December 06, 2012
Software allies smooth path onto emulator
Customers of the HPA/3000 emulator will be watching to see which software companies want to collaborate with Stromasys, to make sure this source of modern, updated MPE/iX servers on Linux iron gets into 3000 shops.
The first HP 3000 manager to take an emulator into production moved the services of very old iron onto a very new MPE/iX platform. IS Manager Warren Dawson’s company was using a Series 947 server which was more than 20 years old to take care of mission-critical operations.
Nearly all of Dawson's third party vendors have come on board and made efforts to ensure their software works. “One was a little slow in doing so, so we made a workaround," he said, "and I made it a permanent workaround. I didn’t know when they would come on board. They came on just before we went live, and we’d already decided to move away from their product.”
In the case of the switch in backup processes, Dawson’s procedures now back up twice as much data, using HP’s standard STORE and RESTORE programs — in less than than when the backup was done using the third party software on the 3000 box.
The change from using HP’s native iron to emulation has also reinvigorated some of Dawson’s MPE software vendors.“I’ve even gotten better support from some of our vendors now that we’re emulating. They see that there’s an extended life in the system, and so a couple of them have made efforts in that regard. We’ve been paying support for years, and for some software we’d hadn’t asked for support in 10 years. They’ve come back to our requests to help us and been very good about it.
One backup software solution didn’t make the transition from 3000 hardware and storage devices to the emulated system. DAT tapes presented an extra effort. Dawson used a utility to copy the tapes to disk, “and for some reason when I did that, it didn’t work properly in the backup software. There was some sort of SCSI issue which was at Stromasys’s end, and they’ve since resolved that issue. But the backup vendor said initially they weren’t supporting the emulator, so we worked something else out.
The Quiz reporting tool from Cognos is part of the software set that’s made the step onto the emulator. The company buys and maintains its Cognos licenses through a reseller, and that partner has handed the relicensing of Quiz onto the emulator. “I haven’t dealt directly with Cognos for a long time,” Dawson said.
Minisoft’s ODBC drivers run on the emulated system, since part of the application’s project is to extract data. Since the databases and the application have been emulated, Dawson’s remains able to use Visual Basic programs, using the ODBC drivers, to do reports as well as updates. However, the complexity of moving toward Windows 7 has raised issues that Minisoft has been helping to solve. Dawson singled out the company as taking extra time to help make the emulation succeed.
“Minisoft’s been the most helpful, because that reporting system started out being the most troublesome. We’ve been having a VB 6 program issue, where those programs ran under Windows XP but are an issue under Windows 7. These programs were written 10 years ago, and the people who wrote them are long since gone. They explained how I could run their software in different ways, with the old driver under VB 6 on XP versus a new driver for .NET on Windows 7.”
Crucial to the process was the support from the creators of the Charon HPA/3000. “I’ve been very pleased with Stromasys themselves, because any issue I raised, they fixed it as fast as they could,” he said. “I was unaware that any of my issues had already been raise, because they way they treated me was that anything I said was important to them, even if they’d heard it before. Not saying they’d already heard it, but saying, “Tell me about this, and let’s go through it.”
The biggest issue I had which I was asking them about was the security key that’s got to be connected into the host PC. We were planning disaster recovery, but we didn’t have a key to replace that one. They said they’d give us a temporary key we could use for a number of hours. They’d already addressed that possibility.”
Support issues with timezone differences haven’t presented a problem. One key Stromasys engineer bridges the gap between Australia’s workday and the one in the US headquaters, Dawson said. The Stromasys labs include operations in Moscow as well as North Carolina.
Get e-mail notice when the NewsWire blog gets a new entry. Just say "Blog Me" in a message to firstname.lastname@example.org.
December 05, 2012
First production emulator wins IT's respect
The first HP 3000 manager to take an emulator into production moved the services of very old iron onto a very new MPE/iX platform. IS Manager Warren Dawson’s company was using a Series 947 server which was more than 20 years old to take care of mission-critical operations. That 3000 had 112 MB of memory. Now it’s working on the HPA/3000 Charon emulator with 2 GB of memory. “We’ve really increased our speed, our memory and our disk,” Dawson said.
“I was testing the emulator over the last 10 months, and I was most impressed with the speed gains,” he said. The gains on month-end processes on the emulated 3000 system slashed the time from almost 10 hours to 65 minutes. “That was phenomenal, and it was on the main database. The guys at Stromasys were very pleased to hear some of the statistics I was churning out. They could emulate, but couldn’t have someone hit it every day, and hit it hard.”
“The users are very happy. They’ve notice their reports are coming up a lot quicker. Instead of 15-20 minutes, in a few minutes it’s done. Performance gains are bigger in some areas than others. The lowest performance gain I’ve found is in backup itself.”Justifying the cost of the emulator became simpler because the HP 3000’s disks kept failing on a regular basis. The HPA/3000 eliminated the difficulty of replacing that type of hardware.
“Because you’re not dealing with physical devices, it’s now made it a lot easier to consider even expanding what we have," Dawson said. "We had a failure of the HP 3000 box every one or two years, and it’s been really hard to source parts here in Australia. The last failure we had was an LDEV 2 disk, and so that became a SCSI disk with an adapter.”
In another instance, an internal cable for a tape drive failed. Parts supply remains an issue throughout the country, Dawson said, since there are few 3000s still running there. But he added that the company searched around the world for that cable. “The best we could get was a two-week wait for it, and we could not wait two weeks for something as critical as that.”
Over the years of moving drives in and out, the cable was pinched and then broken, and “we could not source another cable. We ended up making our own.”
The company has turned off its HP 3000 production machine. “In the end, we had the confidence to do that,” Dawson said. We’ve gone to modern hardware we can get at the drop of a hat, We can almost go into the shop and and saying that one and that one, and one in blue. It removes the need for having specialized spare parts.”
Emulation created a new range of storage space. The company had a project to split its database, due to legal requirements. To do the split, they needed to duplicate the database, and “we wouldn’t have had the space to do that on the Series 947’s disks.”
VMware hosts the virtualized partition where the HP 3000’s emulation resides. “We’ve taken the Stromasys software and moved it to its own VMware environment. It’s by itself, so nothing will impact it there. It’s running really smooth.”
November 28, 2012
As Itanium speeds up, sites fly to Windows
Within the next week, HP's going to ship a new generation of Itanium-based servers. Using the Poulson chipset known as the Itanium 9500, these blade-based systems are going to outperform the current generation of Integrity servers by a factor of 3.29, according to HP.
The engineering gains are impressive. HP tested the new Integrity blades that use the 9500 series against the Itanium 9300-powered servers. Blades start at $6,490 for the 9500-based systems. "For those remaining committed to Itanium and its attendant OS platforms, notably HP-UX, this is unmitigated good news," said Forrester's analyst Richard Fichera. HP's building these new servers exclusively in Singapore, so it can offer three times the computing speed at about the same price.
But even with all that improvement, HP needed to remind the market that these gains were also heading to its Intel x86 Xeon systems. The reason for that reminder: more of HP's customers, such as those leaving the 3000 in migrations, are moving to Windows.
We're not hearing nearly as many reports of migrations which landed on HP-UX systems. The latest news arrived today from Bob Thorpe of National Wine and Spirits. At the Detroit-area IT center, this 3000 pro turned migrator said their customized system is being moved, COBOL and all, to Windows.
"We are in process of having our in-house designed app (using COBOL, IMAGE, and VIEW) converted to NetCOBOL," he said. "We will migrate to a Windows Server platform by March or April next year."
It doesn't matter so much that it took NWS 12 years to leave MPE/iX. What seems more meaningful is that in spite of the Itanium speed-ups, HP couldn't lock NWS into its single-vendor, OS-plus-Itanium environment during those dozen years.The newest Itanium muscle will arrive a little more than two years after HP's 9300-generation Integrity boxes rolled out to customers. These newer blades consume 21 percent less power, led by a new entry-level server, the Energy Star-certified Integrity rx2800 i4.
But dropping the cost of ownership for Itanium has mostly been a pleasure for the existing HP-UX customer. Oracle cast a year's worth of doubt over the chip's future until the courts made the vendor cease, and pledge to support HP-UX and the other operating systems which rely on Itanium. That's one reason HP reminds the market about Itanium's advances and where the improvements will end up: Xeon systems.
With advancements in availability and reliability, HP’s mission-critical Converged Infrastructure will continue to enhance established HP Integrity platforms supporting HP-UX, HP NonStop and OpenVMS operating systems. Over time, these advancements will cascade to mission-critical x86 platforms delivering a single, unified infrastructure for Unix, Windows Server and Linux environments.
That means this "i4" line of Itanium-9500, with its new server blades of a two-socket BL860c i4, the four-socket BL870c i4, and the eight-socket BL890c i4 -- all of these are simply pilot units for the inevitable transfer away from Itanium. How inevitable depends on the customer's trajectory. Windows-bound sites like NWS don't much care how much Itanium can outperform Xeon.
At TechWeek Europe, one writer there interviewed the European head of HP's Integrity business. The website's Peter Judge didn't hear HP expecting to sway many new customers.
According to VP of Business Critical Systems for EMEA Mark Payne, customers still see plenty of performance benefits in the Itanium platform, and would not move across until the x86 platform can match that. Itanium-based systems like Integrity have better mission-critical performance, and users won’t move away until, at the very least, x86 can equal that, said HP.
Unix systems are obviously changing their role in the datacentre, and no one at HP actually suggested they would start to win back business against x86 servers. However, there was a clear expectation that the end of the Oracle lawsuit and the new chips would unlock demand from uncertain customers.
Judge compared the Unix vs. mainframe battles to the future facing the installed HP-UX base. "When we hear that the Unix ecosystem is doomed, we should take some perspective, and expect a similar process to occur. There seems every reason to expect Unix to last as long as the mainframes it failed to dislodge."
HP's message off its own Itanium website shows that it considers "legacy systems" to be its own older Integrity servers. A business case study of manufacturer Steelcase started with the company's use of the Tru64 OS and PA-RISC, then movement to Superdome Integrity. HP seems just as enthused about seeing fewer Oracle licenses needed in the more powerful configuration.
Itanium once had a clear power disadvantage against the PA-RISC chips that drove the ultimate HP generation of 3000s. It took as many as three years for Itanium to catch PA-RISC after the Intel-based systems began to ship. Somewhere in the future of HP's migration campaign, customers like NWS will be hearing more about Xeon systems than Itanium servers. Windows Server, not the Integrity server, is luring migrations.
November 16, 2012
Running a Freeware Emulator: Just Ducky
Editor's Note: I asked several HP 3000 veterans to see how well the installation of the new freeware version of the Charon HPA/3000 emulator worked for them. In yesterday's article, Alan Yeo of ScreenJet led us through a weekend-long journey to get the right VMware and a 2GB Player-ready file onto a server, rather than a desktop. A genuine HP 3000 played a key role. Now with an ISL> prompt on his screen, Yeo plunges forward.
By Alan Yeo
Second of two parts
Okay, so with no documentation at hand (as of last weekend), let’s try ISL>START NORECOVERY
This starts the MPE launch, I get prompted for date and time which I correct, and it continues with a normal 7.5 launch, right the way through to starting JINETD and logging on as OPERATOR.SYS.
You know what they say. "If it looks like a Duck and quacks like a Duck, it’s probably a Duck," and this thing looks like an HP 3000 and would have probably quacked like one if it could.
As far as I can tell I'm sitting at the console of an HP 3000! I’m running in a Putty Terminal, so I'm not going to be able to do any block mode stuff, but it’s good enough to run a whole load of MPE commands and have a look at the created environment. Yes, it still quacks!I don't want to try doing too much perched on my stool in front of a rack in the computer room, so can I access this thing from our network? Immediate answer, is No. It is configured with some strange IP address, so I need to reconfigure it for our network. On an HP 3000 easy just go into NMMGR, but that's in block mode and I'm connected via Putty.
Looking around the screen I see another icon, which turns out to be for xhpterm (a nearly usable HP Terminal Emulator). I launch it, up pops a colon prompt and I logon as Manager Sys. So far so good, let’s try NMMGR; it loads and runs and I do some basic network configuration, validate and exit — and darn I have lost my connection as the IP address has changed. Now what do I do? as I don't seem to have any way to change the IP address that xhpterm is using, and my Putty window has disappeared somewhere.
Let’s try connecting from a real terminal; nope no luck, looks like I have broken this, maybe this demo version only works with its fixed IP? Anyway back to the i7, and decide that I'll shut down the VM and maybe reload. It may have been me but I couldn't find a way to shut down the VM without saving changes, which I didn't really want to do as I had obviously screwed something. So I saved changes.
I thought maybe I'd have to blow the files away and re-extract the CHARON files again, but I thought, well let’s just launch it again! I did, it went through the boot sequence again, during which I spotted that the new IP I had set had taken effect, and magically when I launched xhpterm again it connected. They must have configured it to use the current IP address of the emulator.
Can I get to it externally via Reflection now? Yes! Okay, now we are "Cooking with Gas." (For those non UK readers you'll have to Google that). File transfer a bunch of stuff, and everything works!
Think I'll finish tidying up in NMMGR, but it won't run from Reflection! Why not? What normally stops NMMGR running? Yep, hptypeahead was turned on, but how — I hadn't done it and it’s not a default. A quick search shows that this box has a whole bunch of SYSTEM UDCs set including:
setvar hpsysname 'CHARON-DEMO'
setvar tz 'PST8PDT'
setvar hptypeahead true
Now fine and dandy if I had actually been in Pacific Time, and if I had wanted hptypeahead set (I NEVER have hptypeahead set!).
Bit of a cleanup job to get rid of UDCs and replace with a set from one of our HP 3000s. Driving an HP 3000 with someone else's UDCs is rather like walking around in someone else's oversize boots. They are still boots, they keep the water out, but it just feels a bit uncomfortable, and you can't run!
I do a bunch of file transfers and restores, some COBOL and Transact compiles, restored a database, ran some programs, everything worked. And to be honest I didn't expect it not to!
For those of you thinking of trying the emulator, don't waste your time trying to find something in MPE that doesn't work properly, or a program that gives different results, You won't. I know this sounds too good to be true, but it isn't.
I was fortunate enough to have Mike Marxmeier explain to me a year ago how a hardware emulator works, and basically if you can get the OS to boot, it’s a done deal and anything that runs on that OS hasn't the faintest idea that the hardware has changed. And this is the real MPE we are booting, not an emulated MPE.
The only thing that is emulated is the hardware, so the only place where there might be problems would be in handling peripherals, or possibly the interpretation of error codes from them. Believe me, way beyond my capabilities or desire to go investigating.
So we now have a virtualised MPE 7.5 HP 3000 running on an Intel i7 server (which we have called "Sharon"). It only permits two concurrent users (hey, this is the free version) and I'd defy most people to logon and know that it wasn't a real HP 3000.
I don't know what the final hobbyist version of the CHARON-HPA 3000 package will look like, as I was just being used as a guinea pig tester by Ron. However, this 7.5 box came with all the subsystems I needed to do anything I wanted. If the final hobbyist version doesn't, then unless you already have a 7.5 box with an MPE license then it will be virtually useless to you.
CHARON-HPA 3000 is exclusively 7.5, so you won't be able to take subsystems of your aging 6.0/6.5 9x7/9x8 and use them. My opinion is that for the Hobbyist Licensed version this shouldn't be a problem, as it’s restricted to two users so it’s not like HP would be opening the floodgates on the use of unlicensed subsystems. What’s more, anyone moving from an earlier version of MPE already has a licensed version of them anyway. However, HP is a strange company these days, so I guess we just wait and see what happens.
Commercially, I'm sorry it works, as it will give people more excuses to homestead instead of using ScreenJet's software to migrate. Personally, I like it, as it sticks two fingers up in the air at HP and says "see, if you had wanted to keep all those HP 3000 customers you lost it was technically possible.” And who knows — as ScreenJet's Transact and VPlus migration products also run on MPE, and we now have a new MPE platform, maybe there may be emulator customers interested in advanced versions of Transact or VPlus with all the bugs fixed. And versions that are far more capable than the original HP versions, and are supported!
November 15, 2012
Installing the Emulator: Ahoy, the Disruptor
Editor's Note: As soon as the freeware personal edition of the Stromasys 3000 emulator went live for downloading, I sent the FTP links to several HP 3000 veterans to see how well the installation worked for them. Before we'd follow through on helping to host this freeware, I wanted to see the state of the packaging. Allegro's Gavin Scott also installed it at our request, and his report appears in the forthcoming 3000 NewsWire print issue.
By Alan Yeo
I'm not sure why I agreed to Install the Stromasys CHARON-HPA/3000 freeware. It's disruptive technology to the HP 3000 migration business that my company depends upon. However, as I have spent most of my working life using an HP 3000, it would be nice to always have one available after all the old hardware dies or becomes uneconomic to keep alive.
This is almost one of those stories that went nowhere. There seemed so many stupid obstacles to overcome that I almost gave up a few times -- and that was mainly down to lack of documentation that could have saved hours of work. There was also the fact that instead of wanting an emulated HP 3000 on my desktop, I wanted one on a server where a few of us could test drive it.
Hopefully, the lack of documentation last weekend will have been resolved by the time you try the freeware. But here, over today and tomorrow's articles, is the tale of getting my HP 3000 Emulator into the delivery room and smacking its little bottom until the first little colon prompt appeared.
Part 1: Getting things downloaded and installed, starting with a compatible VMware Player and a 2GB Stromasys file.
My only documentation for this was an email from Ron Seybold at the Newswire, with a link to a 2GB download on the Stromasys site.
- Intel i7/i5 or Xeon CPU with SSE4.1 support; 2 GHz minimum, 3GHz or above recommended.
- 8 GB RAM minimum.
- dDsk space - 0.1 TB + space required to keep HP3K disk images.
- 20 GB is the minimum requirement for the freeware package.
- Two Ethernet ports.
This is the full 2GB VMware kit, uncompress and open with VMware Player. (And an FTP link followed)
CHARON HPPA runs under any of three supported 64-bit Linux Desktop distributions.
Ubuntu 11.10 is our recommended Linux distribution, and is available at no cost. Ubuntu 11.04 is also supported; versions 12.04 and 12.10 can also be used for testing. Fedora 16 Desktop Edition (64-bit). Fedora is available at no cost. Fedora 15 and 17 are also supported. Red Hat 6.2 (64-bit) is available at www.redhat.com; it is a commercial distribution.
Fortunately we have an Intel i7 server that already has Ubuntu 11.10 Desktop as the host OS. Unfortunately it has Virtualbox installed not VMware, and there were a number of horror stories on the net about running VMware and Virtualbox on the same host. This it turned out was not true, however your mileage may vary.
So the first problem was getting the 2GB download. I don't have fast broadband, and to be honest I didn't see the "GB" and read it as "MB" (as who the hell downloads 2GB?) so it was a bit of a surprise when I browsed to the ftp location and started the download and was told it was going to take 23 hours! I think I looked at the screen for a few minutes just to let it sink in that it did say 23 hours and wasn't going to change its mind, it didn't and I killed the process.
The next day with the weekend looming I thought okay, I'll start the download to my PC in the evening and pick it up the following evening (if the connection has managed to stay up that long). This time it told me that it was only going to take just over five hours (don't know what had happened in the intervening day) but five hours meant I was able to check before bed, and as the download completed, plan to do some work on it the following day.
Saturday: The Second Shoot of the season, and me and the dogs were out after Pheasant and Duck, so "Sharon" was going to have to wait. Evening, glass of wine, let’s take a look at where to get this VMware Player thing. Find the VMware site, find the latest version downloads, Oh blast, another 200MB download. Ah I know, I'll logon to the i7 server and download it direct. Strange, if I went to the website from my PC with IE I was offered the downloads. If I browsed there from the i7 with Firefox I got the page, but no downloads offered. Since it is evening I can't be assed to find out why, so resort to downloading the correct Linux version for x86-64 to my PC, and will pick up the following day.
Sunday: Really nice sunny day, unlike the crap we have had for weeks, did I want to spend time working indoors? No, but if I didn't I might never get back to it. Fortunately as it transpired everything took so long and was so broken that all I had to do was wander back and check on progress every hour or so.
Okay I have this 2GB download I need to move from my PC to the i7 server. Easy I'll cut a CD, Windows refuses to copy the file! CD burning software refuses, nay, won't even show me the file to select! That's okay, I'll FTP it. Windows FTP won't even show me the file with a DIR let alone let me PUT it anywhere!
How do I move a 2GB file from my PC to the i7 server if I can't FTP it and can't burn a CD or DVD of it? I could try playing with my PC to see if I could share its drive and do an FTP GET from the i7, but life's too short. I then thought, I wonder what Reflection thinks of the file? Sure enough, it’s happy to show it, maybe it will transfer it? Where to? I need something with PCLink installed.
Ah what about an HP 3000? No problem, Reflection starts transferring the 2GB file (in Binary format Streams) to one of the HP 3000s. It says it’s going to take a fair while even over a 10MB link, but the sun is shining and I can wait.
Becoming a VMware Player
Okay, let’s get this VMware Player set up on the i7 so it's ready and waiting. 200MB is easy to move via a memory stick. Got the file on the i7, follow instruction to right click and open with gedit, it’s a shell script file that it says will do everything for me, including extracting and installing VMware Player if it isn't already installed. Off we go, it has to process the file but shouldn't take long —this is an 3.4Ghz i7 quad core with 8GB RAM.
Time passes. Time passes. Look at the bottom line of the screen: it says it is processing line 450827 and counting, Time passes, the count is on 600 thousand and something! How many lines could there be in a 200MB file? Time passes, I wander back about 30 minutes later it’s on line one million one hundred thousand and something, and as I'm watching it pops up a box to say its finished, but it has an error with some of the characters, do I want to continue? in which case the result will probably be bad! Or do I want it to try a different character set encoding to translate the characters? Okay, says I, "have a go." Oh blast, it’s gone back to line one and started processing the whole frigging file again! Yep running just as slow, time to get outside and do some real work and come back in another hour and a half.
I notice that the Reflection transfer of the 2GB file has finished to the HP 3000, so now I need to get it from there to the i7. No problem: open an FTP connection from the i7 and get the file (binary) leave it running, go get that sunshine.
Pop back a couple of hours later. I have a nice 2GB "Sharon" HPA/3000 file on the i7, and the VMware Player extract has finished! But has the same error! Okay, so on the latest 200MB VMware 5 something is broken and won't install. Give up, or get a long spoon and ask the evil Goggle Empire what it might know.
Okay, lots of horror stories about getting VMware Player 5 running (or rather not) on Ubuntu 11.10. But quite a few people having success with the older Player 4.5 version. Browse to the links from the i7, and this time I can see the version 4 downloads! Select the latest, slightly smaller, and let the download run. Come back, it’s finished, and this time the file has a .bundle extension, so I click and run and it unpacks and installs like a dream,
What next? Okay, find the now-installed VMware Player, and run it, up it comes. But what do I do next (remember, I have NO documentation). It must have something to do with that 2GB file that has a GZ2 extension, so let’s try opening it. Right click, get offered an open by something, which I do and then get an "extract" option. Okay, in for a pound, in for a penny, so off we go, everything unpacks clean as a nut.
Okay, there must be something I select from VMware Player. I click open and browse to the directory where everything "Sharon" had unpacked, and it showed me a single file that it obviously thought it could use. I select it and click open, and wait. I get a warning that something is trying to open Ethernet1 in "promiscuous mode" but that it has been denied, and that if I want to read all about it, a web link was supplied.
At this point I have been playing fast and loose with a bunch of software, so I don't give a damn about something else being a bit promiscuous. Wow! I get a "Sharon" screen, and then a Putty terminal window opens in a bright green. (I start thinking Putty, that ain't going to do Block Mode) but low and behold in the Putty window I see an HP 3000 going through a recognisable boot sequence ending with an ISL> prompt.
November 05, 2012
Accepting Irregular Statistics
We're on the eve of the US national elections today, so a lot of stories are being told about statistics. In many segments of the country, one-third of the registered voters have already cast ballots. We are told that statistically there are under 1 percent of the voters who remain undecided.
A small percentage might continue to matter. And the trends often do matter statistically. For example, Microsoft's Windows XP still represents about half of the PCs still in use, according to metrics company Net Applications. And just this week, the number of Mac users who are clinging to three-year-old Snow Leopard Mac OS still leads the installed base.
And maybe just as surprising, some large and well-known companies are still continuing to embrace their HP 3000s. It's irregular to believe that major corporations continue to use an operating system this dated. Well, maybe not so dated. MPE/iX got its last security patches in 2008, just a little bit farther back than Snow Leopard was created. Maybe because of their stability, both Snow Leopard and MPE/iX continue to serve in the market. One place we discovered this morning is PC Mall, an online sales outlet selling computers that will run Snow Leopard and Windows XP. And they're doing it off software written for MPE/iX.PC Mall is providing an irregular statistic, but they also prop up a trend. The adoption of non-MPE/iX platforms by the installed base has slowed to a crawl. Migration suppliers all predict that 2012 will one of the least active migration years since, well, the 3000 transition era started in 2002.
What's more, PC Mall isn't a complete outlier. Unisource, a $5 billion company, continues to run its operations on HP 3000s.
Both of these pieces of information come by way of the LinkedIn's HP 3000 Community Group. There's 538 of us in that group, numbers that start to approach the membership of the 3000-L newsgroup. Except you can see and connect with every LinkedIn member. New members come on, like those from PC Mall and Unisource, every week. Chris Enderle of Unisource checked in when he signed up.
I still work at Unisource based out of Atlanta and we are running strong on the HP 3000. Unbelievable that we are still running a $5B company on the 3000, but like I tell our CIO, as long as we keep electricity to them, they will chug on forever. We have very bright people writing code, and they do some amazing things compared to when I wrote code.
Code from bright people is creating interesting statistics about the prospects for our election, too. And in about 36 hours that exciting code will give us results of a hotly-contested election. I hope you've voted already if you're in the US, or that you will do so tomorrow if you haven't. It takes full participation and complete tabulation to get to the point where you can accept irregular statistics for what they are -- part of the greater truth.
October 30, 2012
Personal 3000 iron offered for shipping cost
The HP 3000 emulator is still en route toward its freeware personal version. But in the meantime there's still plenty of equivalent HP-badged iron out there in the marketplace. One spot to look is in the shops of the recently-migrated companies.
Lane Rollins of Boyd Coffee sent us a notice about a pair of Series 918s he's been wanting to move out of his datacenter. (There's a Series 979 on hand that's not going away, even though the company has been migrated for several years.) Rollins was looking for a good home for his rack-mounted Series 918 and a standalone 918. Both of the systems are the same power as the personal version of the HPA/3000 emulator software. The rack-mounted unit had an added benefit of an extra SCSI card as well as HASS storage.
This kind of hardware is still circulating in the community, even if it's got as much cost attached to it as that personal freeware emulator. If you can find something like this out on the market, Steve Suraci of Pivital Solutions -- which still sells 3000s -- says you shouldn't be paying more than shipping. Although his company collects systems like this for their depot parts value, they also keep an eye the shipping costs.
We still take them on at times when it makes sense. Some gear is too far away to make sense. By the time we get a mover out there to collect it all, and get then have it shipped back to us we have more into it then we could ever get back.
But the closer a 918 sits to your own shop, the better value it can be -- so long as it's offered free, plus shipping.
"Someone local may be interested if they could pick it up," Suraci explains. A migrating customer who's holding deactivated Series 918s may have to help out on the costs to ship away. To be frank, this kind of server is a better value to the buyer than the seller. Some kind of pickup fee, even from a scrapper, would usually be part of moving out this lower-end 3000. At Rollins' shop, the offer included an LPQ 500 printer with LAN card, a Printronix-based unit, and a few p405 printers.
"The resale value is just about nothing on this lower-end 3000 gear," Suraci said. "Any of the printers might have some value, because of the fact that they are used in non-3000 shops. Shipping cost would still probably be a concern."
October 26, 2012
Taking Care of Too-Great Expectations
Apple is weathering the woes today of an entity which is managing expectations that are too great. Migrators may be laboring under the expectations of moving too much of an HP 3000 to another platform at one time.
Of course, these are very different times for these subjects. Apple set a record for a single quarter. At $35.9 billion in sales ending Sept 30 -- a boost of 27 percent over last year -- it's on a run rate that can make it a $143 billion company during 2013. People continue to call Apple a consumer company, although millions of its devices are powering the mobile needs of business. You simply cannot sell 40,000 phones and tablets -- a whopping $24 million worth -- in a brief 90 days just on the whims of consumers.
So Apple's on a mobile computing upswing, but not enough for the finance analysts. These experts who predict how much a company will earn guessed a little more than Apple posted. So today's a down day for the stock, just at $593, the first time under $600 since last August. HP used to suffer from such Great Expectations. Today, not so great.
However, the HP 3000 has expectations as well. Not for the growth of the platform or an increase in the revenues from its economy. 3000 expectations run to how much of its databases and applications need to be mined and moved -- and how much can remain on a 3000 in near-line storage, ready for the ultimate extraction.MB Foster walked customers through the benefits and strategies of using its UDA Central software this week. This time out in its fortnightly webinar, the company's founder Birket Foster compared the subject of data migration to the expected needs for such a journey. You don't have to bring everything over, even though UDA Central makes it drag-and-drop easy to do so -- even for databases and servers which have little to nothing to do with HP 3000s.
Foster noted during the webinar that some customers are even purchasing 3000s for the specific reason of putting data onto the equivalent of a railyard siding. Of course, that's a a low-speed track section distinct from a running line or through route such as a main line or branch line or spur. But the sidings might still connect to higher speed sections.
"Among the things we've discovered is that when you go to extract your data, obviously you're reading a lot of data," Foster said. "That has an impact on the amount of CPU cycles and bandwidth being used to help data across to the other machine. You have to make sure you understand the timing of when you do that. It wouldn't be a good idea to do that in the middle of the day." But then came the surprises in expectations: 3000s on some kind of new mission, as well as what you can expect to move.
For that extraction reason, some of our customers have gone and bought a separate 3000 to stage the data. They just move the database. They don't move any of the code. They take that database and use it as a staging area to work with it. On the final extraction, they'll go back to the production database. At least they've got a working area where they're not interfering with day-to-day production. You might be able to come up with a very low-cost HP 3000.
There was more to consider about too-great expectations of migration of data.
Some of our customers have been able to work with us to get a methodology that allows them to move just the last month's records, or the last week's records, at the time of moving between systems. That's because all the rest was already staged. History is just history. As long as you can prove that the totals of all of the above equals the total of what you've moved, there's not a problem. Except in cases where you've got revisionist history, the history shouldn't be changing. If you look at it, about 90 percent of your database of transactions didn't happen in the last week or month.
Using this method, a customer could do a first run of data extraction, make adjustments to the process (item names that might be reserved words, different transfers between datasets), and then take a larger segment of the database and repeat. If a migrator has great expectations of making a complete move of data in one pass, they're overlooking these adjustments.
"We've seen customers where it actually takes two days to move all the data, and it ran into some kind of problem," Foster said. "Then we had to check the logs for details." Naturally, UDA Central has a comprehensive logging capability.
A customer from the UK in the webinar, who's moving off a version of Ecometry to an app using SQL Server, said he'd need to check on his permissions to access that target database. It's also essential to be adjusting the expectations for the time to clean up and route these extractions correctly, Foster said. Then there's the understanding that not everything's got to be migrated.
"Customers don't think of all the issues that there might be during the initial stages," he said about more typical sites. The fact that the UK user was on the call showed some foresight. "It's not until they get deep into the project they realize there might be any problems."
You consider how much data you expect to keep, "not only from the accounting perspective, but also for marketing, merchandising and purchasing," he said. "We suggest people start a migration by looking at how much must happen, and do it early. If they discover during that look they need an extra six months, it's better than learning they don't have any time left at all. Know how long it will take from the beginning of the data dumping to the end."
October 24, 2012
Speeding Along Migration's Silver Linings
Migrations off HP 3000s come in varying degrees of difficulty. One set of choices gives sites a way to move their MPE-based environments with the Fresche Legacy (Speedware) AMXW. Even heading to HP's Unix servers, this kind of project can take more than a year. George Willis of the investment portfolio management house Fayez Sarofim talked about their project that moved COBOL and Powerhouse onto Unix.
Coupled with Eloquence as their database, the migration took around 16 months, "largely due to the volume of code that we used," Willis reports.
We used Speedware to help us migrate our portfolio accounting system, “DataVestor”, to HP's Unix Itanium servers. The Unix server was the best choice for us because we leveraged AMXW to emulate the MPE/iX environment -- so that we could lift and shift our COBOL and Powerhouse code with somewhat minimal changes.
But moving away from older HP 3000 hardware uncovered an advantage. "The silver lining to being pushed off the HP 3000 by HP is that our overnight batch improved three-fold," Willis said. "That means we now have a comfortable recovery window before users log onto the system in the mornings."
October 17, 2012
Changing Engines, or Cars? It Depends
Homesteading customers are looking at the Stromasys emulator product as an interim solution before migrations. Dan Miller, a consultant in the community whose roots go back to using MPE in 1975, helps a client who wants to know if the HPA/3000 will perform in place of a Series 9x8 server.
The customer of Miller's runs their 3000 without HP support, but the site has risk avoidance measures in place.
As insurance, they have an additional redundant HP 3000 system on-site should hardware parts become immediately unavailable; besides HP, there are many third party hardware resellers available to replace or repair their hardware. They are assured of software support, as I am retained on an on-call basis should they run into system or software problems. They can also contact the local HP office or other third party vendors for pay as you go software support.
But Miller noted an unusual profile for the homesteader. The company is running "lights out," which in this case means operating with no IT staff in place, except for the on-call Miller. The arrangement which gives ERP and financial processing to about 35 users has been flawless, "but time marches on," Miller says, "and a future migration is inevitable." Perhaps not nearly as close as it might be, if the emulator meets Miller's definition of viability.
For another aspect to the question, customers will weigh how cost-effective any emulator will be. That's a subject where ScreenJet's Alan Yeo says the costs depend on a customer's comfort with MPE's limits and the success of current applications.Miller points out that his consulting client's operations to support a hardgoods distribution business is running without DP staff.
Although they have knowledgeable mangers to handle the day-to-day operations problems, they have no DP staff! Because of the reliability of the HP hardware and the demonstrated success of their custom software, they have been able to avoid the cost of hiring a System Manager, Programmers, or Operators – they run “lights out” and are extremely satisfied with the results of these executed plans.
Needless to say, they are skeptical and hesitant to move away from their current system.
When is that kind of migration due to arrive? Yeo says the same choice befalls an owner of a comfy and reliable car which needs to be certified for a long trip. Buy a new engine, or spend the money on a down payment on a new car? Even at a reduced level of licensing, HPA/3000 is still going to cost in the realm of five figures. Yeo examines the question, "Is an emulator cost-effective?"
If you are keeping HP 3000 hardware running for a couple of thousand bucks a year, then No. If you are reliant on your HP 3000 applications, are happy with the constraints of MPE, and think that will hold true for a few years, then Yes.
The analogy is rather like running an old car that you like and feel comfortable with as your daily transport. It's inexpensive to keep running; the local parts store and local garage seem to be able to supply and fix everything for a few hundred bucks when it breaks. But you have a nagging doubt that if you decided to undertake a cross-continent journey that the engine might just expire with a loud bang. Now someone comes along and offers you a brand-new engine for $5,000 -- what do you do?
1) Decline the offer as too expensive, and carry on as now.
2) Accept the offer, and continue driving your comfortable old car for a number of years, confident that it's got a good engine that's going to be reliable.
3) Decide that if you're going to spend $5,000, that you may as well use it as a deposit and buy the new vehicle that you would really like, one that will better suit your needs for the next decade.
Which is the best most cost effective choice? Ah well, only you can know that, for as with so many things the answer is "It Depends."
October 11, 2012
COBOL You Know, vs. COBOL You Don't
Migrations are in play all over the world between HP 3000 systems and Linux environments. Nobody seems to be reporting very many at the moment, but the Little OpenSource Environment That Could is a regular replacement when a 3000's futures go a-wanting.
All well and good, in many instances. Hiring Linux help is never an issue, but the know-how and replacements for the rest of the 3000 ecosystem are more complex. For example, a customer who's been using scripts in their HP 3000 ops needs a replacement. MB Foster's created one for Windows in UDAXpress, which the company has been demonstrating this year.
COBOL, however, becomes an element that might be integrated tighter than you'd imagine in a 3000 program suite. For example, one recent migration project we heard about included a 4GL-to-4GL Powerhouse.
The decision was made to move the application largely as is, to Powerhouse on Linux, and to Oracle. Porting Powerhouse is not too onerous; apart from a few limitations and differences, you just port the code across and recompile it with Oracle as the target database, and off you go.
There was one catch, and it might become one in a migration near you. Some core calculations can be enshrined in a set of COBOL routines. Maybe they were too complex to write in Powerhouse. So at this point, a Linux-bound customer is looking seriously for a COBOL replacement. They can reach for commercial products which run on Linux, or look to the open source community at OpenCOBOL. Some such migrations are moving from a COBOL they know, to a COBOL they don't. The commercial COBOLs have support staff and training. Open, not so much, unless a third party gets involved.A seasoned migration engineer on an adept team says that OpenCOBOL and Linux had to be blended without help from the OpenCOBOL online forum. This typical sort of knowledge repository for open source "seems to have been read-only, for newcomers at least, ever since last January." When nobody's posting to a help forum, any questions had better be the same as they ones already answered.
OpenCOBOL is open source code that has a commercial counterpart, just like Red Hat commercialized Linux. You can download COBOL-IT to get started with this blend. Freshe Legacy, the former Speedware, was drawing attention to COBOL-IT during 2011.
But 64-bit OpenCOBOL, running on RedHat Enterprise Linux 5.3, eventually assumed the core calcuations which the 3000's COBOL once did. The calculations were in surround code. Sometimes Powerhouse is an application's surround code, but sometimes it's COBOL.
The 3000's COBOL can be compiled on OpenCOBOL 1.1. (Actually running it against a database like Oracle is another matter. There's the calls to HP's intrinsics, plus the exchange of data with IMAGE, to rewrite into Linux intrinsics and Oracle calls.) But there's also a thorough pre-requisite to simplifying the COBOL from the 3000.
1.Remove all the code relating to long-dead product ranges that would never be purchased again. Good policy in all migration cases. Your migrations should well begin by studying all the programs that need not be migrated, because the end-users don't use them anymore.
2. Make the code almost completely ANSI-compliant, using COBOL's functions for date calculations instead of any home-grown ones. The 3000 COBOL's ENTRY points are already simple enough. They might be a lot of trouble to code around, and OpenCOBOL supports them anyway.
The blend of OpenCOBOL and Powerhouse works very differently than the 3000's, which requires this bit of technical refitting: keeping the OpenCOBOL on 64-bit.
A 32-bit OpenCOBOL is needed if Powerhouse, itself 32-bit, is to call COBOL subroutines. So you do a COBOL wrapper for the subroutine, which makes it possible for Powerhouse to RUN it as a separate executable -- and 64-bit OpenCOBOL will be okay, now passing and returning the variables in a file.
If all of the above sounds like the effort of home-grown application development from the 1980s and 1990s -- workarounds galore -- of course it is. These migrations are moving applications that were constructed during that era. It might also serve as a leg in the journey moving the COBOL you know to a COBOL you don't on Linux. Especially if that COBOL is open sourced. The reputation that Linux bears -- being a hands-on environment -- survives, especially powered by reports like this.
Database: vendor-supported. Environment: same. COBOL: Perhaps best chosen as a commercial tool with support. And be vigilant about the run-time costs, which never existed under HP's COBOL II. That was the final COBOL, by the way, that HP ever created, using a well-honed languages lab. By the time COBOL became important to Unix or Linux, HP had left the compiler business to third-party experts.
October 01, 2012
When HP's SUSAN Won't Say Enough
Emulator vendor Stromasys has sold a few instances of its Charon HPA/3000 virtualization engine. But there's even more interest in the free version of the product. Not much surprise there, considering the average budget for a company that's sustaining its 3000 in production use.
However, there's another kind of 3000 user who's looking at this personal freeware. Developers of MPE/iX code -- mostly consultants, and some tool and utility providers -- are expressing an interest in downloading the freeware version. When they do this, they'll require some strategy to tell their other software that the emulator is actually an HP 3000 with a valid HPCPUNAME and HPSUSAN number.
The HPCPUNAME and HPSUSAN combination is used by third party vendors to validate a license. If the combo doesn't match, most software won't load at all. (At best, you might get a demo version, but that's more of a personal computer fallback.) Stromasys is looking at the issue for the freeware it calls the A200. The paid version of the product uses a USB stick with these numbers encoded, which makes any Intel i7 Core PC capable of running a utility like Adager or an application like Ecometry.
But the days of that HPSUSAN being a unique number -- identifying only one MPE/iX licensee -- are over. CEO Rene Woc of Adager said that as HP began to use and re-install these numbers, creating its own 3000s out of HP 9000 servers, duplicates have emerged. But the combo of HPCPUNAME and HPSUSAN is still needed for verification. Even it it's not unique, it's still not generic.
"Way back when, with an HPSUSAN I would be able to tell you the HP 3000 model, and even the serial number, I believe," Woc said. "Today that's not true anymore. It's not a unique number."
This leads the users interested in freeware HPA/3000 to a challenge which Stromasys must master: How to give hundreds of freeware emulator users a way to employ their valid HPSUSAN numbers with third-party software. Only using the full complement of software on the emulator constitutes a complete test, Woc said.Stromasys founder and chairman Robert Boers is examining this challenge. He ran one concept past me last week.
In the commercial version of the virtual HP 3000, the HPSUSAN is located in the license key (set to the number the customer specifies). The freeware A200-sized emulator has no license key. We can program into the code a fixed number (such as 123456.) Would that work for non-commercial use?
Commercial or not, a fixed HPSUSAN won't verify third party software which expects a number registered with the vendor. While a fixed number would satisfy MPE/iX so it could boot up on the virtualized 3000, it doesn't seem likely that it would meet the validation requirements which utilities, development tools and even some applications require.
And if you're going to test it, you cannot ignore the third-party software.
A lack of that kind validation might render the A200 HPA/3000 -- Stromasys calls it equivalent to the power of a Series 918 -- nothing more than a proof-of-concept demo.
"If you're going to test it, you cannot ignore the third party software," Woc said. "At some point Stromasys might offer a freeware version where they charge a nominal fee for the administration of producing a USB device with some HPSUSAN and HPCPUNAME that would work with the third party applications and software. The third parties would have to update their license algorithm," Roc said while he considered the challenge, in order to use a generic number. To avoid triggering upgrade fees during the testing, that HPCPUNAME would be the lowest possible tier of 3000. A Series 925 comes to mind, probably the slowest 3000 ever released into the market.
But HPCPUNAME is half the value for those algorithms. And even a user who has valid HPSUSAN numbers will need to deploy them on a personal freeware A200 model of the emulator. For example, Taylor Lumpkin of the Hire Experience consultancy continues to develop for Ecometry e-commerce users. Hire Experience was founded by key employees who designed and built the Ecometry app suite. Lumpkin loves the idea of skipping the use of more 3000 hardware with an emulator -- just like his company has skipped Windows hardware by virtualizing the OS on Macs.
"We continue to develop for Ecometry on MPE/iX," he said, "and being able to run it on one of our existing i7 Apple machines, right along all of our Windows virtual machines, would be of great benefit. We could deploy machines to our remote developers and eliminate the need for connectivity."
We are still a HP Partner, and HP have allowed us to have free MPE for over a decade now. We also own a small pile of 918s which all have legitimate HPSUSAN numbers with the HP license converted into our name by HP -- back when they still did that.
We love virtual machines and have not had to run any hardware on Windows now for seven years. This has proven to be a huge resource saver -- as we have eliminated all downtime which used to accumulate to 7-8 person days annually, all by running our desktop and server hardware on OS X exclusively.
Intel i7 distinctions aside -- the only i7s referenced by Stromasys so far have been PC hardware running Linux -- the profile of a developer customer usually matches lowest-speed units. Developers rarely need the commercial-grade, production-level horsepower of 3000s to cut and maintain code. But a free version of a 3000 might get in the way of a Stromasys sale.
Simply put, using the HPA/3000 freeware as a development tool would only benefit the developer. Stromasys seems to want to introduce the A200 product into the end user customer base -- a group of users who would likely need a paid version of the software to put the emulator into production.
If the HPSUSAN licensing challenge could be solved, the A200 could become the realization of the mythical Series 908. That was a model of 3000 which HP was going to sell to its developers in the 1990s for as little as an equivalent PC development system. The 908 was much more of a programmer's wish than a genuine HP product. But it illustrated just how little budget was available to development teams for 3000s.
Instead of the Series 908, HP introduced the Series 918DX. The server was only available to members of HP's DSPP developer program. Each came loaded with all of HP's subsystem software. But it was sold by the vendor that created the 3000, so each 918DX had an HPSUSAN which could be registered with any third party for software validation. A few third parties included their software with the 918DX. Most saw the low-power system as a prospective sale, instead of way to expand their installed base through a reference or a proof of concept. The emulator, being novel technology that appears to be a marvel to much of the market, could use proof of its concept.
September 25, 2012
App design changes induce homesteading
At the e-commerce site Musical Fulfillment, ERP manager Chris McCartney would prefer to remain on the HP 3000 with an app which has been working well. But migration mandates at these kinds of satisfied sites are triggered by many things, including the loss of HP support or acquisitions by larger companies.
Musical Fulfillment has used the Ecometry application for more than 10 years. The company even moved up to the N-Class HP 3000s just a few years ago. "We were hoping to get a few more years out of it before we had to make a decision to upgrade or move to a different ERP system," McCartney said. Her firm is the parent company for musical suppliers such as American Music Supply.
"Personally I love HP 3000s," McCartney said. "They are sturdy, they run forever, and they are just one box with none of this load balancing across multiple servers."
The alternative path away from the 3000 induces changes, sometimes ranging beyond a new environment. It starts with a new vendor, in McCartney’s case. Red Prairie acquired Ecometry’s creators last year.
Accomodating a new vendor for your app induces change, but a shift in the product's designs through a replacement version can be more serious. It might be difficult to duplicate all software functions with a replacement package, even the Unix and Windows-based replacement from Red Prairie.
Customers want to carry their business rules and customized code to another platform. It can be tough if application changes have drifted away from the MPE designs. More than five years has elapsed since the vendor last cut Ecometry code crafted for MPE/iX.
One question to put to a vendor inducing a migration to a commodity replacement version: What have you got in software that's going to emulate the operations of our last MPE/iX release, for a minimal amount of change to my site's configuration? If you don’t like the answer, getting more years out of the 3000 is another plan.
Making a stand on MPE, or the 3000 hardware, poses a different challenge to delay a migration. One way to go is some kind of emulation, to get MPE/iX onto newer, non-HP 3000 hardware. AMXW, an emulated platform, lets a company move their MPE environment to a Windows, Linux or Unix host. The product creates a shell above the host hardware.
Another ploy to stay with an MPE application that's working well might be to deploy the MPUX software, from Ordina Denkart. (It's sold as part of the company's ViaNova 3000 solution.) MPUX is really meant for a move to HP-UX. MPUX hasn't been moved forward to Windows 7 yet, and there's no mention of Windows 2008 on the Ordina Denkart website. Additionally, it would not be a good fit for an Ecometry client.
"Most Ecometry sites go to Windows," said MB Foster's Chris Whitehead. "Therefore, MPUX is not a fit. Secondly, most Ecometry sites want to convert Suprtool, COBOL and JCL to native scripts."
Newer e-commerce solutions which run on other hardware platforms have licensing practices in place with application vendors. For something like the Stromasys HPA/3000 emulator, an Ecometry Direct Commerce customer would need a license from Red Prairie to run Ecometry on it. This licensing for emulator is an area where the app vendors are waiting for customers who want to go to a 3000 hardware emulator.
Open Systems Ecometry looks to be the path of least change for an MPE Ecometry customer. Whatever ROI there might be could come from eaving the 3000 community (and its need for MPE IT skills) behind. At this point in the assessment, customers ought to ask themselves if they could hire someone to replace their MPE/iX experience.
However, the adoption of SQL Server or Oracle represents another significant change in using Open Systems Ecometry, plus new exposure to reporting tools which are unlike Cognos Quiz.
At least Suprtool — also key surround code for Ecometry — runs on HP's Unix. Many more Ecometry sites have moved to Windows when they stayed with Ecometry. Suprtool is also within reach of running on Windows, because Robelle has made tech changes enable a new version for Linux, too. Some sites remain on watch for a more sensible package to replace Ecometry on MPE.
September 24, 2012
E-commerce sites examine migration plans
One year after the Red Prairie buyout of Ecometry's owner Escalate, the e-commerce suite is getting a more secure open systems future. Ecometry once represented the largest and most vital part of 3000 growth, especially during the dot-com boom of the late 1990s. A list of 3000 customers circa 2003 showed that one customer in five was using the e-commerce software.
When satisfied users of 3000 apps are sparked to make a migration away from the server, they often rely on the considerations of their longtime app vendor. At e-commerce and catalog firm Musical Fulfillment, manager Chris McCartney is still searching for a solution that will improve on the 3000-based Ecometry software she's managing. Her company serves several e-commerce sites such as electricguitar.com.
Even though the Ecometry app's supplier Red Prairie sells a commodity version of the software, that migration target is not registering a higher note at McCartney's company.
"Unfortunately there is very little ROI in an upgrade to the Open Systems Ecometry," McCartney said, "so that is a hard sell."
The application and its creators have moved from part of the Escalate software group to an even less prominent part of Red Prairie, which now has 43 software solutions in its stable after a series of acquisitions during the last three years. But recent signs point to protection for this application suite -- at least its versions built for non-3000 environments.Red Prairie has now found some gusto in keeping the Ecometry solutions in its portfolio, according to Birket Foster. His MB Foster company partners with Red Prairie to help migrate Ecometry sites like Musical Fulfillment. Companies still looking for the right moment to make their migration cover a wide scope.
"They're either really, really small, or really, really large," Foster said. A tough economy stalled smaller customers, while the larger ones might be hemmed in by corporate IT procedures. An app that's doing its job won't always be put in flight by a vendor's support changes. Costs are always a factor, whether the customer is large or small.
"You pay Red Prairie to migrate your application," Foster said, "then pay someone else to migrate your surround code." Migrating off Red Prairie products might cost $5 million, while shifting to an open Ecometry version could be $2 million, he estimated.
McCartney has been researching alternatives to the MPE-based Ecometry software called Direct Commerce. The alternatives don't improve the application capabilities at the company which fulfills orders from American Musical Supply, zZounds.com, ElectricGuitar.com, and SameDayMusic.com. A competitor, Musician's Friend, made a migration off Ecometry to Junction Solutions several years ago.
"We have been looking at other solutions like Junction Solutions," McCartney said, "but I have yet to find a good reason or some fantastic functionality that the others provide that we don't already do with Ecometry/Direct Commerce. Maybe I am missing something."
If there's a silver lining in the situation for the Ecometry sites still running a 3000, it's the renewal of vows on the part of Red Prairie. There's a version that runs under Windows, and another that runs under Unix. But Red Prairie is no longer in danger of running off Ecometry sites to one of the other software packages in its sea of ships, Foster said.
"There was a question of whether they would keep Ecometry in the boat," he said. "That was a year ago, though. The management shuffles are done."
Just as MB Foster has hundreds and hundreds of customers who continue to rely on its products for MPE/iX, McCartney would prefer to remain on the HP 3000 with an app which works well.
The company has used the application for more than 10 years and moved up to the N-Class HP 3000s a few years ago. "We were hoping to get a few more years out of it before we had to make a decision to upgrade or move to a different ERP system," McCartney said. "Personally I love HP 3000s – they are sturdy, they run forever, and they are just one box with none of this load balancing on multiple servers."
September 06, 2012
Core memories spark a cold start for 3000s
Editor’s Note: Jon Diercks, the author of the only comprehensive MPE/iX administration book, offered us this story of the 3000’s very first year. It was a time of HP retreat from the minicomputer market: HP staff resigning, others unselling a system touted just months earlier as “a happening,” as the slogans of 1972-73 said in HP labs and offices.
Diercks worked at Anderson University in the 1990s alongside Tom Harbron, who’d been the college’s computer department director during 3000’s first months on the market. Diercks said Harbron was heavily involved in early discussions with HP about MPE and IMAGE.
The institution began as Anderson College, and its very first HP 3000 was one of the earliest models. Diercks said the bragging line in those days was "Anderson College has the first HP 3000 ever installed anywhere between the Rockies and the Appalachians."
Harbron’s report on the 3000’s 1973 is part of Diercks’ 3000 memories, and so he’s contributed the writing as part of our 3000 Memoir Project — in all of its authentic, human and humbling beginnings. It's the first story I've read that details the 3000's retreat. An HP employee who couldn't look his customers in the eye about the 3000, and so resigned. A man whose job was to unsell the 3000s -- and later would bundle the greatest software HP ever wrote, IMAGE, to the Classic hardware, which not long after, fell behind the state of the art.
By Tom Harbron
Reports of problems with the HP 3000 operating system, MPE, continued to be received in the opening weeks of 1973. While it was not encouraging, I had confidence in the basic soundness of the 3000’s design and the integrity of Hewlett-Packard to ultimately deliver what had been promised.
HP’s Phil Oliver called and scheduled a meeting with me for February 6, 1973. He brought along Bob Stringer, who had replaced Ed Pulsifer as the District Sales Manager; Ed McCracken, who was now HP's Market Manager for Government, Education, and Medical Markets; and Jay Craig, who was a new HP salesman from Indianapolis. McCracken would tell me, years later when he was the 3000 division manager, that the morning in my office was the most difficult day of his career. The people that HP hired were, mostly, an honorable group of people.
On that day in 1973, they had some bad news to deliver. Specifically there were seven points:
1. HP cannot bring the software components of the system up to full specifications before Fall 1973.
2. They are devoting “maximum resources” to correcting the problem.
3. The system will currently support no more than 4-6 simultaneous users.
4. HP will loan an additional 64K bytes of core storage to bring the system up to this 4-6 user level of performance. (We had ordered the system with 64K bytes of core storage.)
5. IMAGE will be further delayed to January 1974.
6. Because of hardware difficulties, a slower console printer would be provided.
7. They would like us to cancel the contract. Lacking that, they wish to amend the contract.
It was a tense meeting. McCracken was going about the country, visiting customers, and unselling the 3000. It was hard for everyone. Phil Oliver would return to his office later that day and resign. He told me he couldn’t look his customers in the eye. Ed Pulsifer had already resigned for similar reasons.
I resisted their pleas to cancel the contract for four fundamental reasons.
First, I had faith in the basic design of the system. I had run benchmarks that had come in almost exactly where I had predicted from the timings in the ERS. MPE was clearly a better design than nearly anything else then available. The combination was more cost effective than anything else by a factor of two or three. Moreover, I had met many of the people involved with the project and had confidence that they could do the job, given the time and resources.
Second, there really were no viable alternatives on the market at the time. DEC tried long and hard to sell us, but in the end their salesman conceded that the systems DEC had were either too feeble or far too large for our needs; the HP 3000 was a perfect fit. IBM tried hard to sell us on various timesharing patches to their systems, none of which worked well. The only systems available that would do the work were the XDS Sigmas and they cost five times as much as the 3000.
Third, I thought that HP had to make the 3000 succeed if they were to remain a growth company. At that time, HP had about half of the instrument market and could not significantly expand their market share without anti-trust problems. Their other market areas, such as microwave, were respectable, but in small markets that were not growing very fast.
The only way that HP could continue to grow at historic rates was by entering the computer business in a major way. With $25 million already invested in the 3000, they were unlikely to write off that investment and content themselves with their existing markets. If the 3000 failed, they would have had to immediately start over on another computer project.
Fourth, we had already invested several man-years in application development for the HP 3000 at the time that HP was trying to unsell the system. It would have been a financial disaster for us to write off all of that work and begin again with a different system. We really had no option beyond the 3000.
Years later, in a speech before the Users’ Group, McCracken said “I want to thank those of you who had faith that the HP 3000 would succeed at a time when many at HP had profound doubts.” I’m sure he was thinking of that cold February day he spent in my office.
August 17, 2012
NorCal transit will run its 3000 route again
Anybody who wonders where HP 3000s are hanging on can grab a rider strap on the Alameda-Contra Costa transit service. The public entity AC Transit just opened up a one-year contract to maintain its two HP 3000s, along with the applications.
The systems under maintenance are a Series 957 and a Series 987. If you're scoring at home, these are servers built and sold during the 1990s -- still powering a California organization with duties to ferry 191,000 riders daily with a fleet of 584 buses. The District’s service area extends from western Contra Costa County to southern Alameda County, and the organization employs 1,863 employees.
As if that's not enough, this contract -- which is out for bids until Tuesday, Aug. 28 at 10 AM -- has a provision for extension. The district isn't sure when it will be able to stop using those 9x7s.
There are many 9x7s -- well, more than you'd think -- still serving the public or working in private firms. Keeping these two servers online 24x7, including holidays, is going to demand that a winning bidder can prove they know the 3000 and MPE. Things being the way they are in California municipal government, this contract starts its period of performance three days before the bids are due. We'd bet they're going to approve a support supplier pretty quickly.
At the sole discretion of the District, the contract may be extended up to 12 additional months in increments of three months. This is to accommodate the uncertain end date for the District’s use of these HP 3000 computers.
The minimum requirements sound like they could be from any HP four-hour response contract.
- The Contractor will provide a toll-free telephone number, staffed during typical local business hours, to allow one or more AC Transit contacts to report all service requests.
- The Contractor will provide unlimited telephone consultations for both 957SX and 987SX systems during AC Transit business hours.
- The Contractor is responsible for all parts, labor, travel, testing equipment and phone consultations on covered equipment and necessary on-site visits.
- The Contractor shall provide staff to make at least one weekly on-site visit to manage the tape backups and perform a physical systems health check.
- The Contractor must provide staff that can physically attend to these systems and operate independently 24x7x365 with or without a District Information Services escort once an appropriate District security badge has been provided.
- The Contractor is required to demonstrate proficiency in HP 3000 system support by compiling a checklist for AC Transit review and approval. The selected contractor will employ this checklist daily to determine and report on the hardware and software health of both systems.
- The Contractor must have access to a local facility that stocks good-order hardware equipment in a location such that needed parts have been tested and are delivered and installed within four business hours of the incident opening for priority 1 and priority 2 incidents.
Whoever takes on the maintenance "will create and employ a 24 hour/7 day automatic alert system that will alert them to a hardware failure on the Series 957SX and/or the Series 987SX." That's far from commonplace. There are plenty of businesses that don't have auto-alert failure systems.
All this by console “via a dedicated workstation with a serial connection to each of the two HP 3000 computers.” It doesn’t sound like networked remote console access, but that wasn’t a big part of the 9x7 experience anyway. IT at AC Transit calls the shots on application maintenance requests.
It's not a static system, by the looks of the bid. Item 7 calls for a contractor to "Perform ‘software application support’ allowing I.S. staff to make program or application changes. These changes would require that the affected application have no user access while the changes are being made."
There are 37 pages of this RFP, but some of them even do a good job of outlining the base competencies for any 3000's management. Like the 30-item list at left (click for details). It'll be interesting to see what AC Transit eventually replaces those 9x7s with -- and when -- considering that it's got such a customized app suite on its hands. At least there's 12 months of work here, and probably more, for a persistent bidder. Got to keep the buses running in the public's interest. Yeah, buses: as fundamental a transport technology as a Series 9x7.
August 16, 2012
Moving Data in Migrations: the Tools, and Who Uses and Develops Them
Arby's sandwich chain turned off some HP 3000s recently, but moving its data stocked a menu's worth of practices and tools. Based on a report from Paul Edwards, the journey worked smoothest when expertise could be outsourced or tapped.
Edwards described part of the project as a move to Oracle's databases, facilitated by Robelle's Suprtool and Speedware's software. The former supplier has retained its name for 35 years by now. The latter has become Fresche Legacy, but DBMotion as well as AMXW software is still available for data transfers. In the photo at left, the veteran Edwards is in motion himself, flying on a 1968-69 US Navy tour on the USS Hornet. He figures he's been working with 3000s half his life, which would give him enough time in to witness Robelle's entry into the market, as well as the transformation of Infocentre into Speedware, and then to Fresche Legacy.
I'm standing on the right. The two young guys kneeling down are the enlisted operators that ride in the back of the plane. The guy standing on the left is our Crew 13 Aircraft Commander. The aircraft is an S-2E Tracker Carrier Based Anti-Submarine Warfare Navy aircraft. It has a large propeller attached to a 1500hp Wright R1820-82 engine -- one of two on the plane.
Some of the data moves at Arby's went to Oracle, he reports. "They were using Oracle for part of their operations. Using Speedware with Oracle was interesting. Most of that was dumping data with Suprtool or Speedware, then formatting it in the layout they wanted." Suprtool has been guided and developed by Neil Armstrong at Robelle for nearly two decades. He recently marked his 20th year with the vendor, according to the Robelle newsletter.
Arby's also took its payroll application off the 3000, "and it went off to a service bureau. We had the file layouts that bureau wanted, and so it was a lot easier. We just said, 'this field is the one on the HP system, and this field on your layouts is equivalent.' We just matched them all up. We had some where we could say 'forget about that field, we won't need it.' "
But the transition to Oracle, as performed by a team that was supposed to be experienced in the database, was not so easy.The Oracle contractors "had absolutely no clue about how to do migrations," Edwards said. "They'd never done any before."
The migration of data from a well-polished, longtime set of 3000 applications is just as crucial as moving code, selecting a replacement app, or testing what's been moved. And it's not as easy as it might seem to find contractors who've done a migration, especially any who know MPE. Plenty of systems from other vendors haven't been worth the time to migrate. The HP 3000, with its lengthy lifespan, often sports apps that are decades old. Almost as entrenched as Armstrong has been at Robelle.
The avid racing cyclist this summer completed 20 years' worth of "helping to make Qedit and Suprtool great products," Robelle reported in its newsletter.
Neil worked at one of our customers in Ontario, then worked for us in British Colombia, then worked for us in Alberta. At one point Neil moved to Anguilla in the Caribbean to work on Robelle software with Bob Green, our president. Lastly, he moved back to Canada and works on Suprtool and Qedit near Niagara Falls. He is currently our Software Architect, chief systems programmer and a big help for difficult technical support questions.
During his time in Anguilla, Armstrong raced in the 2004 John T Memorial Bike Race. The photo at left shows him with Bob Green cheering him on at the finish. Armstrong has been quick to the pedals for as long as I've known him; as a fellow cyclist, he rides at a rate I can only dream about. But his work in Suprtool -- especially in recent years getting it to Linux, and soon to Windows -- must have been as steady and careful as a rider navigating a busy, two-lane, no shoulders road. That's a tool that began its life in the 1970s, when Edwards was still in the Navy Reserve and working at HP as an SE. Imagine what's been changed in Suprtool over those decades to get it to Suprtool Open.
Sometimes great care to advance a product unveils its rewards when it's compared to other migration methods. It helps if you can call on some military precision during critical transits, too. At Arby's, Edwards and the IT staff seemed to be glad Suprtool was on the migration menu.
August 06, 2012
What You Need to Do and Check for SLTs
At a recent visit to a customer's shop, VEsoft's Vladimir Volokh spread the word about System Load Tapes. The SLTs are a crucial component to making serious backups of HP 3000s. Vladimir saw a commonplace habit at the shop: Skipping the reading of the advice they'd received.
"I don't know exactly what to do about my SLT," the manager told him. "HP built my first one using a CD. Do I need that CD?"
His answer was no, because HP was only using the most stable media to build that 3000's first SLT. But Vladimir had a question in reply. Do you read the NewsWire? "Yes, I get it in my email, and my mailbox," she said. But like other tech resources, ours hadn't been consulted to advise on such procedures, even though we'd run an article about 10 days ago covering CSLTs. That tape's rules are the same as for SLTs. Create one each time something changes in your configuration for your 3000.
Other managers figure they'd better be creating an SLT with every backup. Not needed, but there's one step that gets skipped in the process.I always say, "Do and Check," Vladimir reports. The checking of your SLT for an error-free tape can be done with the 3000's included utilities. The venerable TELESUP account, which HP deployed to help its support engineers, has CHECKSLT to run and do the checking.
There's also the VSTORE command of MPE/iX to employ in 3000 checking. If your MPE references come from Google searches instead of reading your NewsWires, you might find it a bit harder to locate HP's documentation for VSTORE. You won't find what you'd expect in a 7.5 manual. HP introduced VSTORE in MPE/iX 5.0, so that edition of the manual is where its details reside for your illumination.
It's also standard practice to include VSTORE in every backup job's command process.
There's another kind of manager wouldn't be doing SLTs. That's the one who knows how, but doesn't do the maintenance. You can't make this kind of administrator do their job, not any more than you can make a subscriber read an article. There's lots to be gained by learning skills that keep that 3000 stable and available, even in the event of a disk crash. Management might not respect the 3000's ability to take on new developments. But a company always respects the 3000's reliability.
CHECKSLT, and care and feeding of SLTs, are well-covered in a NewsWire column written by John Burke almost 13 years ago. His advice still holds today.
HP’s documentation tells us we need to have a current SLT. And that it can be created using the TAPE command within SYSGEN. If you look hard enough you will also find the warning that the CSLT you may have created when doing an update or manage patch is not adequate. That is about it for SLT recommendations.
Is this recommendation correct? Well, in the sense that it is necessary to have an SLT created by the TAPE command, then, yes, it is correct. You can re-install your system in the event you lose a drive in the system volume set using this SLT and appropriate other backups. But is this recommendation complete? I think not.
As has been proven countless times, the people who write manuals (and not just at HP) are not out in the real world. They are not running shops where if you get a six-hour maintenance window once a month you consider yourself lucky. They are not running shops where you have to have procedures that can be understood and followed by someone with only basic training in system operation. They are not running shops where cell phones go off like July 4th fireworks as soon as anything unusual happens.
You can find HP's VSTORE page in that 5.0 command manual online, just like the NewsWire's advice. Vladimir, you find him in your office, if he's traveling your way. But managers also find that he recommends our advice -- perhaps because we first get the instructions to do it, and then have our reports checked. Do and Check are words to live by, not just for managing 3000s.
August 01, 2012
Just how good were those good old days?
NewsWire subscribers who receive our email updates have heard that I'm collecting stories about the early 3000's days. I'm working on an autobiography of the 3000, written "as told to" me, by the system. I've fielded phone calls and gotten some nice email stories. Today's was great fun to read and instructive, too. That's because the negative experiences in our lives are remembered clearer than the positive ones.
What I mean to say is that war stories are more fun to read, chock-a-block with details. Before I offer an excerpt from today's story, I want to make an observation about the 3000's life. It wasn't always the better time we prefer to remember.
Even the president of the Connect user group falls prey to this memory. In his column in the latest user group magazine, Steve Davidek remembered days when HP was packed with people eager to service a 3000 customer. After a disk head crash in 1984, Davidek recounted three HP employees he knew by name who chipped in to resolve the problem. A different time indeed, when Davidek managed just one Series III HP 3000.
Our HP sales rep would visit every month or so just to see how we were doing. Some months he'd even bring a Systems Engineer along to check on things. It was amazing.
Dave Wiseman, who says that "Most of you will know me as the idiot dragging the alligator at the Orlando conference, or maybe as the guy behind Millware," told us a tale of days even earlier in the 3000's life. Buying a system from HP in 1978 meant investing in a terminal to test your application -- before HP would even fill the system order.One of the first three HP 3000 customers in Southern England, Wiseman was managing at an IBM shop looking for a better system. "I called the HP salesman and asked him in," he says.
What HP never knew is that if the project went well, there was a possibility that they would get on the shortlist for our branch scheme – a machine in every UK branch office. That would be 45 machines when the entire UK installed base of HP 3000s was around 10 at the time.
So the salesman came in and I said that I wanted to buy an HP 3000, to which he replied, “Well I’m not sure about that, as we’ve never done your application before. Why don’t you buy a terminal and an acoustic coupler first and make sure that your application works?"
"Okay," I said. "Where do I buy a coupler from?"
"No idea," he replied “but the 2645A terminal is $5,000."
So he bought the terminal, and then tested against HP's 3000 in a UK office. "I started dialing into the Winnersh office. (I still have the telephone number and address engraved in my heart). On occasion when I needed answers, I would drive over there and work on their machines."
Wiseman goes on in his early history to praise the improvement that the 3000 delivered to Commercial Union Assurance.
I recall our durability test was to unscrew the feet on the 50MB disc drive and push it until the disc drive bounced off its HP-IB cable. On more than one occasion the cable came out and you could just plug it back in and carry on working. Try that with an IBM and you could expect two days of work to get it restarted.
The IBM guys couldn’t understand how we could run so many users on such a small box, but we were always looking for improved performance -- as we already had the largest HP 3000 around. There were no tools available in those days, so we used tricks like putting a saucer of milk on each disc to see which one curdled first. (Okay, that's not really true. But we did spend a long time just standing there touching the drives lightly just to see what got hot.) We did a full system unload and reload every three months, and unloaded and reloaded most databases at the same time.
Davidek recalls his warm feeling of having ample HP support, but he does recognize it as a bygone emotion. "The customer experience today is probably not ever going to return to those days, but I would love to come close," he writes. "HP is working on this issue, and with a little luck, we may get there."
War stories are useful for more than the warnings about potential pitfalls. Even from 30 years ago, they remind us the good old days were not as good as we remember. They also remind us how our initiative made the bad times manageable. That's a confidence builder in these uncertain career times.
A 3000 manager needed a little luck, all the way back to the beginning. I'd like to hear about your lucky and unlucky days. Call me at 512-331-0075 if you want to chat, or email me. By recalling both the good and the bad, we can chronicle the middle path for that autobiography.
July 31, 2012
How Support (With)holds Key to Emulation
At some HP 3000 sites, the servers are working to deliver IT services to a subset of customers. One site in Virginia which handles healthcare administration is keeping an N-Class server online using Amisys/3000. But the issue which concerns the company isn't so much the tech capability of the Stromasys HPA/3000 emulator. Licenses with existing vendors worry this prospect.
Cognos/IBM is at the top of the list for this company, even through they dropped Powerhouse support long ago. Powerhouse has been an integral part of the Amisys surround code. Cognos wasn't the friendliest company to negotiate with during the MPE heydays. An emulator license for Powerhouse would have to be arranged with IBM.
Other arrangements would include a license for Amisys itself, which is part of the McKesson Group by now. This is one of the software programs running in Virginia where support is being paid -- Adager is another -- and so there's a better chance of getting that license transferred. A transfer license is what's needed for this market. However, it's a lot harder to start up a transfer discussion after you've terminated support.
There are going to be other kinds of prospects where the Stromasys software has a better chance today. The sites where support is intact, or there's plenty of in-house code with no vendors to negotiate with, will have a smoother path. In a spot like that Virginia IT shop, licenses are linked to vendors which will likely expect some support.
This support renewal raises the price bar for the emulator, of course.We've reported on other strategies on how much an emulator ought to cost. Comparing it to the value of the company works, so long as the 3000 drives the entire company's IT. That's not the case in Virginia. HP's BladeServers drive Windows, while the HP-UX systems are running Facets (another healthcare app), but are soon to be replaced by Sun's hardware and Linux. (Take that, HP; Oracle has won that contest.)
The IT pro we interviewed at the shop asked first about technical capability, and the HPA/3000 has been shining there. In very short order he asked about licenses. Hewlett-Packard's emulator licensing is ensured, and companies like Adager, Robelle and VEsoft have expressed little concern about licensing their software for an emulator. (There are tests to be run at some vendor labs. Robelle reported early this year that the HPA/3000 passed all of its tests through February).
Although it might seem like it's hard to convince a CTO or CIO to maintain support contracts on 3000 software, extending support can keep some options open -- especially if there's no clear migration project timeline. Support relationships, renewed regularly, look like they're holding the key to admit the emulator.
July 24, 2012
Make backups, but a CSLT is just as vital
Many homesteading HP 3000 shops are working with limited system administration. If you're reading this blog, that probably doesn't apply to your own 3000 shop. But you can pass on advice about backing up to any 3000 site you know. A backup of applications and databases isn't enough.
The CSLT needs to be fresh and available, too. The Custom System Load Tape tells the 3000 how the configuration is set up for devices attached to the system that you're restoring. (The original SLT that was distributed from HP has a generic configuration. This customized SLT reflects your physical configuration of your specifically-built system.) Also referred to as a boot tape, it contains the system load utilities, diagnostic subsystems, base system files, and other HP system files such as IMAGE, FCOPY and EDITOR.
A CSLT is generated with the system generator (SYSGEN) utility. You can build a CSLT for individual systems, each with a different configuration, after updates. These configurations tell the 3000 what other volumes are available to accept data. You can also put a full backup on the end of a CSLT, but it's better to have that backup on separate tapes. (Separating a backup from the CSLT also speeds up creating a CSLT.) Consultant Paul Edwards advises that managers make a CSLT at least every other time during a backup, plus having two tape drives on each system. "Being paranoid makes for a good system manager," he says. "If you're not paranoid enough, you better have a good resume."
Overlooking the CSLT is so common that even some admin pros have done it from time to time. For one such pro, an A-Class 3000 was recently rebuilt and had its apps consolidated. But the rebuilt system didn't have its CSLT freshened, which was discovered when the boot volume failed.
We lost LDEV1 in the 'system' volume-set. The apps and databases are fine, but I'd neglected to make a fresh CSLT once the rebuild/configure/setup was complete. Fortunately, all the data volumes are protected with Mirror/iX -- but rebuilding the system volume accounts, network config, administration jobs and so on has been a pain.
An honest mistake like this is not one you need to make yourself. Even if, as another 3000 consultant notes, your shop has gone into Frugal Mode while it makes in-house moves. You have the right to be wrong in Frugal Mode. But you really don't want that right, unless you've got plenty of extra time."This is not a mistake I'll ever make again," said our CSLT rebuilder. "Until it's complete and stable, I now make a new CSLT at the end of each day's changes (in addition to the regular backups we do as Store-To-Disk, and then FTP to a NAS). Hopefully I'll never have to go through this hassle again."
AUTOINST updates temporary copies of the system libraries and then creates a CSLT. This can take up to two hours. One set of instructions to do this is available in HP's MPE/iX 7.5 Maintenance Manual. The process hasn't changed much from the 6.0 release of MPE/iX, either. We've got a pointer to each manual; just search for "CSLT" in each PDF file.
We've previously published a paper by Gilles Schipper on the use of BULDACCT versus the STORE command while backing up. The ;DIRECTORY and ;ONVS= options are a key there. What's more, one of our most prodigious contributors Brian Edminster is sketching out a technical, but easy-to-use, Automation of Backups and Effectively Eliminating Use of Tapes article.
Do all the backup that you need when protecting that HP 3000. It's a computer that established the gold standard for recovery in case of powerfail situations. But the success of its backups falls in the lap of the system's managers or owners. (Running CHECKSLT from the TELESUP account will verify if a CSLT is still good enough to boot your system. You'll want that CSLT to run on any tape drive, not just the one it runs on. Alignment issues kick DDS drives out of service regularly.) Don't forget about keeping your CSLT healthy.
July 18, 2012
Disbursing 3000 Gear to the Next Life
Some 3000 shops which have made migrations are looking for a new home for older equipment. At the Boyd Coffee Company -- whose motto is "Fuel, not fashion" -- director of IT Lane Rollins has systems which once did good work but might be ready for retirement. Or repurposing.
"I need to get some clutter out of here," he said, "and I'm not sure who is reselling HP 3000 gear these days, or if anyone would be interested in the stuff. We’ve been off of the 3000 for five years now."
His list includes two Series 918s, including a Jamaica storage unit (HASS) and an extra SCSI card. Also on hand at the Coffee Company's shop: DTC controllers, an HP line printer with a LAN card, plus terminals.
If the gear doesn't seem like it could launch a thousand ships, it could well be equipment to keep a 3000 ship afloat. This kind of inventory won't draw much resale value, but some service companies need it for their spare parts depot. There's nothing quite as good as replacing a failed bit of hardware with an identical unit.If your 3000 gear simply needs a fresh home outside of your datacenter, contacting the support firms which do business in depot repair is a good place to start.
Pivital Solutions' Steve Suraci made a point of this kind of resource earlier this year. A manager announced a need for a 3000-model HP router to keep a storage unit linked up to a 3000. A search across the Web -- rather than a text message to their service company -- eventually solved the problem. Suraci said that's not an acceptable service level.
That provider was willing to take this company's money, without even being able to provide reasonable assurance that they had replacement parts in a depot somewhere in the event of failure. There are still reputable support providers out there. Your provider should not be afraid to answer tough questions about their ability to deliver on an SLA.
Other companies with a need for 3000 parts might be found on our sponsor list: Genisys, the MPE Support Group, or The Support Group. Sometimes a migrated manager simply needs a volunteer to accept a 3000 shipment, or perhaps defray the cost of getting the older gear onto a UPS truck -- or even into the back of a pickup, if the support vendor is nearby.
Boyd's is completely migrated, but there's one last piece of 3000 gear that's not going anywhere yet.
"Someday the 979 will go away, but need to keep it for the time being," Rollins said. We'd bet that some historical archiving is at the heart of that hardware's duty. If you're interested in the Boyd's gear, you can contact Rollins via email or at 503.907.2555.
July 17, 2012
Find SSD on SCSI? It might be time for 3000
SCSI remains the primary method to connect disk to HP 3000s. That means that most of the Solid State Disk (SSD) memory-based devices won't serve for MPE/iX storage. But it's not impossible to make the quantum leap from rotational to RAM storage. It might be worth the experimentation, given the upsides. In short, if a manager can find the SCSI, honey, find the time -- to experiment.
Starting at about $250, the devices are not costly anymore. And even in the more-rare SCSI units, 120-480 GB models are available in online stores.
"You would need to find a SCSI SSD," said Larry Kaufman, a systems engineer with BayPointe Technology. "They are out there, but there are not a lot of choices. Here is something to consider: When a SSD fails, the failure is likely to be catastrophic, with total data loss. HDDs can fail in this way too, but often give warning that they are failing, allowing much or all of their data to be recovered."
Kaufman offered a note about the old HP's "silverback" disks used with 3000s. "They would make tons of noise for weeks, sometimes months, before they would die."
That sound of rotation is also the sound of slower operation, of course. Once you go non-rotational, you won't go back, said one consultant. Using SSD tech in a 3000 carries the usual warning that was often quoted in the 1990s and onward: This is a storage unit not tested for MPE/iX. But there's plenty of tests not yet performed for SSDs. Like any advanced technology, SSD has also got some emerging downsides now being discovered in the field.
For example, there's the lifespan of an SSD unit, which at its heart is just a array of RAM devices. To be fair, some HP 3000s are running with rotational storage that was built before Y2K. SSDs are still new enough, and have tech advancing so quickly, that an MTBF for them is matter of belief in marketing rather than tests.
Arrays help extend these Mean Time Between Failures and offer hot-swap. The 3000's array choices run to the XP and VA models from HP. It can be surprising to see that some well-schooled 3000 managers haven't explored array choices yet. Everybody has something to learn, whether it's RAID or IPv6.
There's also some experimentation going on with setting up a redundant array of SSD units. One tech white paper reminds users that the TRIM command for controlling SSDs should be supported in order to array SSD units, and that older operating systems usually don't TRIM. And by older, the paper's author means Windows XP, which was last released about the time MPE/iX was getting its last disc design. No TRIM in MPE/iX, of course.
The question for the typical 3000 manager is whether SSDs are more reliable or less reliable than hard disc arrays. Choosing XP or VA arrays is a more commonplace storage upgrade for HP 3000 storage configurations. On that subject, Craig Lalley -- who lives to update your 3000 storage -- arrayed the choices.
You could put an XP array with 128GB of cache as a disk subsystem. HP has a whole line of these high speed fault tolerant arrays: XP256, XP512, XP1024, XP12000, XP24000 and a name change to the newest P9500. They all work with MPE. They have software options that provide "hardware mirroring across the WAN", local mirroring and snapshots.
The XP array is much newer, in that they are still making them and the models get updated every other year. Also the XP array is extremely fault tolerant, it is built to be set up, turned on and never turned off.. You can add cache, replace boards without losing data or host connectivity. The VA array only holds 4GB of cache, 2GB per controller.
The XP is several steps above the VA (to put it mildly). You can literally connect hundreds if not thousands of hosts to an XP Array. Comparing an XP array to a VA7410 is like comparing a VW Beetle to a Formula 1 car.
There's a similar speed increase available by moving to SSD. Native support for SSDs on the 3000 was being explored as far back as 2009 by one indie support company. In the current state of the art, the best way to exploit this new storage technology for MPE/iX will be using the Stromasys HPA/3000 emulator. It employs whatever storage is attached to the Intel i7 Core PC running the emulator.
July 16, 2012
Still ample time to study IPv6
A new guidebook to learn the latest-generation Internet protocol will be a useful tool for 3000 pros, at least any who are expanding their networking skills. And while it's true the HP 3000 won't ever support IPv6, this style of trailing the current net technology has been part of the 3000's experience for decades.
In 1984, for example, IBM's System Network Architecture (SNA) ruled a vast swath of a very diverse computing landscape. Almost three decades ago, 75 percent of corporate-level data processing was performed on an IBM or "plug-compatible" system. And about half of those systems were supporting SNA devices. So there you had it: a networking protocol used by more than a third of the world's corporate computers, and completely unsupported by Hewlett-Packard's business computers.
In a story in the venerable InterACT magazine, Sharon Fisher wrote this dominant and rising technology wasn't for sale, just adoption.
The most important fact to remember about SNA is that is is not a tangible product, but an abstract concept. You cannot buy SNA; it is neither hardware nor software. Instead, it is definitions, rules, protocols and formats that govern the structure of hardware and software.
So HP embraced SNA for the 3000, its only business computer, but late. As usual. Adopting standards early might look good in a tech planning presentation, but in practice can be as useful as messaging between a fridge and a TV. We heard as much from a veteran of standards-based networking in the 3000 community.Chris Bartram manages networks for a DC-area company today, practicing so far forward he's a field expert on HP Site Scope, an enterprise monitoring tool that lets you do a variety of agentless monitoring tasks such as web page checks, database queries, and TCP port checks. But he made his name delivering 3000-ready networked products through the 1990s and onward. Netmail/3000 was his best-known product, email for the 3000 at a time when proprietary HP Deskmanager mail systems dominated the MPE landscape. Netmail was built upon standards like MIME, technology far more complex than a corporate-controlled standard such as SNA.
Bartram, who since 1995 has managed a server in Virginia that's hosted 3000 NewsWire web pages, checked in with us about IPv6. Like 3000 net guru Jeff Kell before him, Bartram doesn't see support of IPv6 as much of a practical payoff for most companies. "I’m fairly up on IPv6 in general," he said, "though I don’t plan to implement it anytime soon. There’s just no real incentive yet."
Some of the big guys have finally started pushing it into the real world, but since I’m operational and accessible on IPv4 – and many of the tools and some of the OS network stacks really haven’t seen that much public exposure yet -- so I’m not ready to take on that extra overhead. Perhaps if I was just launching an Internet presence and my ISP was only granting IPv6 addresses (as Cox or someone like that is doing) then it might force my hand. But for the moment the hackers probably have a better handle on all things (and vulnerabilities) IPv6 than I have time to keep up with, so I’ll hold out a while longer. IPv6 gets blocked completely at my firewall.
Since we’re never going to see a working IPv6 stack on any of the 3000s, we’re going to be trapped forever in the “compatibility zone” anyway. But that’s okay – I don’t think I’ll ever feel the need for my 3000s to be able to chat with my refrigerator or microwave or light bulbs.
Bartram refers to the IPv6 promise of enough IP addresses that anything smart enough to network can be given its own address. It could be a good while longer before IPv6 takes a practical role in IT management anywhere but at Internet service providers. But it's probably not a bad thing to study if you're seeking fresh employment, though. O'Reilly has a fresh book on the subject that includes references as current as Windows 8 and Windows Server 2012.
It's entirely possible that only Linux will be fully engineered to understand and adopt IPv6 by the time the next-gen addressing standard becomes practical technology. But it's become a different computing world that it was in the '80s and '90s, when HP routinely played catch-up to the more dominant vendors' standards. That was the typical HP tech strategy, even in those fastest-growth days of the 3000. However, it was an era when being first could be on par with being the best implementer. "HP wasn't always in the forefront," consultant Olav Kappert said while he reviewed his 3000 history with us last week. "But when it came out and it was done by HP, it was done right."
July 11, 2012
Web console resets, environment rebuilds, dumping form printers lead Hidden Value
I switched from an A400 to an A500 some time back, and I only realized I had not set up the remote web console after the console was down. Where can I configure this? This last time my only access was via VPN, or verbally over the phone. ("What can you see? Okay, so type...") I want to be able fix this myself next time. The console is the built-in one, and not an external box.
Gilles Schipper replies
You can configure your web console from the main console via the GSP interface. Specifically, the command you're looking for is LC (LAN configuration).
This command can be invoked even while the system is up and running by typing ctl-B (control and B together). For more help, at the GSP interface, type HELP, then HELP LC.
Craig Lalley adds that if you arrive at a password roadblock and need to clear a console back to the default login, "at the physical console, hit the GSP reset in the back of the system, then press P on the keyboard. It will reset the passwords."
I need to rebuild an environment from one HP 3000 system to another. Trouble is, we want to have groups from the same account end up on different user volumes. Is there a way to do this using BULDACCT?
Keven Miller adds
BULDACCT was made for processing complete accounts. Do BULDACCT CHC%VSACCT=MEDADV_1. Then edit BULDJOB1 for the other group, changing MEDADV_1 to _2
After Donna Hofmeister notes that "newacct has to be done on both volumes," Mark Ranft replies
What I usually do is use BULDACCT to move the entire accounting structure. Then I surgically PURGEGROUP and NEWGROUP (with the appropriate HOMEVS= and ONVS= options, plus CAP= and ACCESS= etc) to duplicate the special groups.
Jim Hawkins notes
MPE/iX RESTORE supports changing volume set specifications on RESTORE (vol, volclass volset) . That is, you can take files/groups/accounts originally on one volume set and place them onto another volume set.
We are still using multipart pre-printed forms and RS-232 dot matrix printers, all direct connected to our HP 3000 for our MANMAN purchase orders, invoices and checks. Obviously, these old HP printers are getting tired and problematic. Our immediate target is the printing of purchase orders. Are there ways to migrate from these form-based printers without breaking the bank?
Lisa Christiansen replies
One option that would work without requiring programming expertise is using ByRequest software from Hillary. We are taking the print ques from MANMAN and overlaying the form/jpeg to print the forms. We also have it set up so that the reports (invoices, POs, and electronic payments) can be emailed to the various parties.
Ed Stein adds
We moved to the ability to laser-print, or email purchase orders in PDF format, using eFORMz and eDirect from Minisoft. For vendors who require a faxed purchase order, we email the PDF of the purchase order to an email-to-fax service, otherwise we email them a PDF copy. We only laser-print purchase orders for vendors who either require a mailed copy or do not have an email address or fax number.
Can I configure a disk larger than 36GB on MPE/iX 5.5? The reason the customer is locked into 5.5 is they use Oracle 7.3 — so I'm now trying to set up a test bed to see if I can move the system up to 7.5 and still have all the wheels stay on. But we might not be able to migrate off 5.5, and the only way to get the enough disc space is to try going to 73GB drives. It doesn't look like 73GB are supported, but that doesn't mean someone hasn't been able to make them work
Jim Hawkins replies
Although your mileage may vary, there is nothing "unsafe" about trying to access 73 GB on 5.5. Of course, there is a 4GB cap on LDEV1, regardless of disk's physical capacity). What's unsafe is the fact that you're likely running on 15-plus year old equipment -- and running on disk-to-firmware-to-HBA combinations they were never tested by us at HP.
When researching "Large Disk" support in early 2000s, my observation was that disk size constraints were last re-engineered around the 4GB mark during MPE/XL 2.x (early 1990s). From there, most of the code worked pretty well up until 512GB, where we run off the rails of some 30 bit limits.
What didn't work okay up to 512GB was addressed by the "Large Disk" patches for 7.5. Things like REPORT running out of space to display a sector count, DISCFREE output etc. The only allocation change I made had to do with trying to spread IO across disks in a more granular way -- really an optimization, rather than an inherent limit.
Allegro's Stan Sieler adds
I remember that when Allegro encouraged HP to handle big disks, I found some definite problems in some obscure things — and that's why we filed a number of bug reports (including JAGad48770 in 2000) and then lobbied with Mike Paivinen, Louis Runnestad, and the CSY lab manager at the time to get really big disks supported. (Ironically, that triggered extra work for me... I had to make sure that De-Frag/X supported them, too.)
Unfortunately, I can't recall the details of what things had problems. Some of the things were offline utilities (for example, DISKCOPY, or DISCUTIL's SAVE command) that could be significant. Some were cosmetic (e.g., incorrect output from DISCFREE).
In the back of my mind, I have a suspicion that on some releases MPE might create some disk-based data structures that are too small — in a way that could potentially cause an overwrite of data ... but I really don't remember what it was, or which MPE release (might have been *very* long ago, sorry!).
My old notes say that 5.5 with PowerPatch 5 had an "official" max disk size of 18 GB and that I thought I'd seen a working 36 GB drive on a system. The same notes say that the base release of 5.5 officially had a 9 GB max disk drive size.
June 28, 2012
Long MPE future, longer list to learn about
Up on a favorite technical mailing list of mine, one HP 3000 manager laid out his future for MPE and the 3000. While it may well be a long one, he's now looking to learn IBM mainframe technology. Yes, studying up for work on a system whose legacy is even longer than MPE's.
Yes, I know that COBOL is dead and there's no future in mainframes. Somehow, I think they're both going to be around as long as I am. Remember, I work for state government. We have at least one agencywhere their idea of application modernization is converting from Assembler to COBOL.
Our veteran manager wants to make a shift that looks like this:
The common element in there is COBOL, a language almost always essential to supporting 3000 applications. In spite of its repeated death announcements, COBOL's in use on every platform being run by enterprises today. Not everywhere, by a long shot. But since COBOL training won't be a big part of our manager's tech learning list, he could move on to newer tech. Maybe you're interested in Java, for example. One of HP's arch rivals is streaming free training for a language that's being classified as legacy. It should be so lucky.
Virtual Developer Day: Java SE 7 and JavaFX 2.0 is being billed as "everything you wanted to know about Java including Java SE 7, Java FX 2.0 and the roadmap to Java SE 8. Discover how Java can help you reduce your project risk and build more advanced, more portable desktop clients." Oracle's set up a webpage that leads you to videos and PDFs of slide sets.
One other 3000 expert, who arrived at MPE after a career in IBM mainframes, said there's lots of crossover between the two platforms. Job control, for example, reads about the same. "JOB cards, EXEC for RUN, FILE = DD in IBM land, etc. -- you could probably take an IBM JCL listing and understand it with little problem."
"As to COBOL, quite similar as well. The CICS [terminal display module] is sort of like VPlus, in that it uses subroutines and program-managed content and buffers. The non-display feature allows you to hide variables on the screen and pass them along which can be handy."
3000 managers may not understand how fortunate they are to have their strong fundamental experience with a classic enterprise environment. Another migration veteran reported the long-lived MVS is just a more wooly version of MPE.
"I suspect that having worked on an HP 3000 is one of the best preparations for working on the IBM OS," he said. "Because my impression of MPE is that it was MVS without the fluff."
June 26, 2012
Reductions via migration feed 3000 pipeline
It's a fact of life in 2012 that HP 3000s are being retired by some companies. Migrations are far enough along -- in some cases -- that backup 3000s are finally being turned off and sold. Many migrators report there's a 3000 still running just in case, even after the MPE applications have been replaced, re-hosted, or re-written.
However, the exits of these servers are usually not from service to the community. By this year, the latest-generation 3000s are coming available on the market once they move out of an IT shop. At the US sandwich chain Arby's, an N-Class 3000 is on offer at a price below $4,000.
Speedware, which is now making a fresh mark as Fresche Legacy, moved out more than 730 HP 3000s between 2002 and 2011. Speedware (the fourth generation language) provided the landing platform for Arby's move off MPE, too. Paul Edwards recently announced an entry-level N-Class for sale that once worked at Arby's. Recently, Fresche reports that it's landed a new project to "eliminate the mainframe environment at a major North American railway, providing $10 million in cost savings and improved performance."
Not all of Arby's change in platforms is spinning off of HP's strategy, however. Arby's and Wendy's married up over the last few years. They're splitting up after a short union, and the IT resources are being dispatched as a result. A 3000 that feels stale to one company may look fresh to another.Edwards reported that a PC-centric, Web-based instance of Speedware was the target for Arby's newer serving of servers.
Several years ago, I was involved in the migration of the Arby's HP 3000 Speedware applications to SpeedWeb on an Intel platform. The system for sale has been used as a historical lookup system by Arby's since then. Now, Wendy's has taken it over and is selling it. Arby's and Wendy's married, and then divorced a while ago.
With new 3000s being limited to such "eliminations," there's a bit of a silver lining in seeing some MPE systems rotated out of companies. People like Edwards, and especially the community's hardware brokers, can offer newer hardware to homesteaders as a result of these migrations. This one is a 440 Mhz single-processor with two internal and 16 external drives. Here's another MPE license that's come online, too.
This silver lining doesn't exist in other legacy modernization projects. Fresche Legacy shared a press release today about the replacement of a mainframe system at that railway. The target for the project is clear: Linux as the OS, along with Sybase as database. It's a commodity solution, something more industry-standard than HP's Itanium/Unix target.
Commodity computing platforms provide high functionality and high performance at a competitive price. The 100-plus applications support multiple business processes in major functional areas. This IT transformation will enhance the railway’s technical infrastructure, reduce costs and improve the efficiency of its IT applications.
A key step toward getting older systems replaced is convincing the migrator of the overall cost savings while spending capital costs on hardware and OS and migration services. Fresche had to "perform a mainframe migration discovery, assessment and analysis."
This analysis provided the customer with a clear understanding of the challenges including; inventory of objects, re-hosting requirements, re-hosting tools, effort, cost and timeline required to accomplish the mainframe migration. Fresche’s extensive migration modernization roadmap convinced the customer that this migration and re-hosting would create significant value by helping the railway more effectively meet its short and long-term business goals.
However, we don't read this eliminated system as an HP 3000. This is a server IBM's probably still selling, which means the used hardware won't be as unique as an N-Class system whose highest bid -- so far -- is $2,500, plus the fees to ship it.
Wendy's never based its decision to move out the 3000 on the server's service to the company. "I don't think Wendy's used it at all," Edwards said. "The migration was finished long before the merger. I assume Arby's used it only rarely. I believe they merged all the Wendy's data onto the [SpeedWeb] Intel system."
So while these two sandwich giants go through the painful period of separating everything -- including IT staff, now on different floors of the same building -- a 3000 has gone out the building's door. Then onto the market as a bargain for some homesteader. Re-hosting doesn't provide refreshes of hardware availability in other markets yet. That's because the AS/400 and mainframe markets don't face hardware needs that cannot be met by system vendors like IBM. Migration feeding the needs of homesteading -- well, that might be a part of the ecosystem which HP didn't understand when it first estimated the freshness of the 3000.
June 22, 2012
Database changes target weekend's quiet
The HP 3000s still serving throughout the world are often on all-day, every day duty. But activity can quiet down on the weekends. That's when Terry Simpkins, Director of Business Systems at the manufacturer Measurement Specialties, wanted to add an index.
"I need to add a new index into a detail dataset," he said. "But I'd like to complete this change on Sunday, if possible, while there are no users on the system." Simpkins hoped that DBGeneral would be able to adjust a Jumbo dataset in IMAGE, but a problem cropped up that didn't have an immediate weekend solution using that Bradmark tool in his IT belt.
The detail is a 'JUMBO'. The new index will be to a new Automatic master set. Using DBGeneral, I get an error when I attempt to 'activate' the change. It says the number of blocks exceeds the IMAGE max. The detail is very large. Is there some undocumented switch that needs to be specified for DBGeneral to work on JUMBO databases?
Whether DBGeneral has any undocumented Jumbo dataset switch didn't come up in users' replies to the issue. But Sunday support expectations, and the built-in nature of the alternative tool to DBGeneral, Adager, rose up. Not even a better tool can enable a change to such a very large detail dataset, however.
"I think that you probably are exceeding the MPE file size on your detail dataset when DBGeneral tries to add the new index pointers," Pete Eggers said. "My choice would be Adager over DBGeneral years ago, and I suspect that would still be true. But even Adager cannot create datasets that exceed MPE's file size limit." But these are the instances when such advice helps teach about the limits which applications must respect in IMAGE/SQL, as well as the tools to work around them.
"If this is true, your only option is to either reduce the record count of the dataset first -- if for instance, you can send some of the records to a history dataset or file," Eggers added. "You might be able to split the detail dataset into two datasets to add the index to the one with the field in question, but that probably won't happen on a Sunday, as all apps that access the dataset will have to be modified."
Steve Cooper pointed out that a Sunday call to Adager's support line would be answered promptly, "even if it is early -- and they will take your order for a copy of Adager, which will have no problem dealing with Jumbo datasets, no hidden switches needed."
Wyell Grunwald, a veteran of more than two decades of 3000 management, echoed the praise for a database manager like Adager. At Measurement Specialites, an empire of 3000s around the world manage manufacturing operations. Grunwald added that the situation there "Sounds like a repack would be in order -- to remove all your logically deleted records."
Keeping a 3000 in production use as a homesteader is most responsible when tools like database managers remain available -- and their support teams are quickly contacted, even on weekend hours. Best of luck to managers who are using this less-critical time to improve and expand the reach of their 3000 data. It's easier with good tools.
June 19, 2012
Keeping up with Cognos can demand a lunch
While IBM's enterprise juggernaut keeps rolling out in front of HP's, the share of business that Powerhouse contributes is miniscule. HP 3000 operations comprised far more of HP's business while the vendor was still selling MPE. Nudging along Powerhouse technology has become a tricky assignment for indie tool providers who need changes. Sometimes an informal lunch works a lot better than any official action on tech agreements between IBM and the long-time 3000 partners who hail from years before IBM bought Cognos.
In the era of this kind of response, 3000 software vendors such as MB Foster have promised Powerhouse-using customers like Dave Vinnedge of Accuride that their software would continue to keep up with Cognos changes. But communications into the 3000 community have become a low priority for the IBM Advanced Development Tools group. (IBM didn't respond to requests for comment on this story.)
Support for Powerhouse at Accuride is $6,500 yearly, a figure that buys the highest level of access for an HP 3000 user: Vintage Support, created in 2009 after the $5 billion merger. It hasn't improved via the association with IBM.
"During disaster recovery testing, about the only time we ever called Cognos for quick support, it took about two hours for them to get us a 'disaster recovery key', " Vinnedge said. "At that time Cognos support needed to contact our Cognos sales rep for an okay. We have not yet tried to contact IBM's support during a DR test." Contact is tough; it took over 15 minutes to find a US support phone number to add to Accuride's DR docs, "and that was from using links in IBM's emails they sent when they merged in the Cognos support."The changes in IBM's response have had an impact on vendors' ability to track methodology across software changes, according to MB Foster's CEO Birket Foster. An informal lunch is the kind of initial contact that's now needed to stay in touch, all to enable the final 8.49F version of Powerhouse can stay in step with Foster's UDALink. In the interim, a customer like Accuride needs to maintain two separate Powerhouse dictionaries – 8.39 and 8.49 – just to keep using third party tools with Powerhouse. Accuride's got an automatic 8.49 DR license, but the 8.39 version it needed to remain in step with UDALink requires special dispensation from IBM, a company not easy for Vinnedge to contact.
"There's no direct jeopardy here," he said, "but we're tired of having to be sure that Powerhouse changes get compiled into two different dictionaries. We do not want to have to call IBM support during a disaster, or a DR test. We'd also like to be able to drop Vintage Support – or at least negotiate a lower cost."
Accuride has been a Powerhouse user so long that they remember when Cognos was called Quasar. A vendor like MB Foster began Cognos relationships in that era, but these partners must now recalibrate their efforts to stay in touch with the much larger IBM organization, Foster said.
CEO Foster said he figured a lunch between engineers was a faster way to get technology exchanged. Within a few weeks, MB Foster added support for the 8.49F version. IBM has stated that 8.49F is the last release for the HP 3000. Along with the update of UDALink, "MB Foster support quickly fixed two bugs we found in a new option," Vinnedge said.
Pinned in the middle of such an exchange is Bob Deskin, a Powerhouse product manager who's been working with the language since 1980. Today he's not able to make any policy statements, even while he posts messages on the Powerhouse newsgroup. "Although Vintage Support does not provide for any development support, it does allow customers with a legal support requirement to continue to use Powerhouse," Deskin stated on a Powerhouse mailing list.
Ken Langendock, a Powerhouse services provider offering new development, integration, conversion and migration, sees indelible value in using a 4GL. But he believes managers have lost the taste for continued investment.
"They seem to think it is a dying language, and resources are getting harder to find," he says, "even though there are still COBOL programs out there churning out data tirelessly. Programmers coming out of school have never heard of a fourth generation language. They would rather work in a 3GL with a Web front end. Give me a common data dictionary any day."
Another Powerhouse developer consultant, Richard Witkopp, says "Few programmers dislike Powerhouse. The hatred comes from upper management. They have to pay for it, and I think that's where Cognos killed its golden goose."
Brian Stephens adds, "I wouldn't hold my breath on IBM doing anything for Powerhouse. They bought Cognos for their Business Intelligence tools. Powerhouse was an established product they could maintain and use to open some new doors."
Without a clear initiative from IBM to revamp the products to support a streamlined platform base – Windows and Linux, Oracle or MySQL or SQL Server, a web look and feel – customers feel like their skills with a proven tool will continue to drop in value.
"The biggest issue is the invisibility of Powerhouse," said Darren Reely, another consultant in the Powerhouse community. "Most IS-type people seem to have never heard of it, and that is reflected in the job market. Besides looking ugly, supporting more platforms, the capability to run on the server, and a need to occasionally trick it, what can a Quick app do that an iPad app can't? The cutting edge companies use the hot new platforms and tools.
"Some things have a slow death. Powerhouse looks to be in that category," he added. "Hey, COBOL is still alive."
June 18, 2012
Powerhouse future spurs 3000 adjustments
After IBM's Cognos acquisition, the Powerhouse language now offers slower development prospects. Some long-term users are adjusting their expectations further in 2012, even while they work to make other software integrate with the fourth generation language.
It might be easy to pin today's prospects for Powerhouse onto IBM. The computing giant purchased the creator of that product, Cognos, in 2007. The years since that purchase have frustrated some users who try to rely on the 4GL. While the concept of a 4GL remains a useful Advanced Development Tool (ADT), the potential for adapting Powerhouse or extending its reach looks challenged.
One example is at the manufacturer Accuride, a 3000 shop using Powerhouse. Dave Vinnedge says he's been working to integrate the latest, 8.49F version of Powerhouse with MB Foster's UDALink software. As Cognos was edging into the realm of IBM, it was changing internals in the Powerhouse data dictionary. These differences were not communicated to such independent software allies in time to make the changes so the products might integrate. More than four years later, the liaison between a now-small Cognos ADT operation to allied companies still isn't working on an official level.
Customers expect this kind of technical exchange to continue between their vendors. But reports from the installed base of Powerhouse customers indicate that IBM's interest in Cognos remains largely in the Cognos Business Intelligence products. The ADT tools like Powerhouse look forlorn in comparison.
It's not like developers using the product have all lost faith. "We had a project to migrate from the HP 3000 a few years ago," says Mike Godsey. "I was told to compare a conversion to Powerhouse Web and a full Java design. We estimated it would take seven developers three years to do the Java, or four programmers two months to move to Powerhouse Web. Powerhouse Web was chosen. We delivered ahead of schedule and zero defects. How many Java apps can claim that?"
But the support from Powerhouse's new owner drives down hopes. "Realistically, what future does PowerHouse have?" asks Vaughn Smith, a developer working in HP's OpenVMS environment with the 4GL. "The products are 30-plus years old, but they still perform quite well on the right hardware. IBM has stated that Powerhouse will remain supported as long as it makes business sense. That said, there are no future updates planned at this time."
On the HP 3000, IBM supports only the 8.49F version of the language, tools which also include Quick, Quiz, and Powerhouse Web. To get an idea of how long ago this version was crafted, the IBM support document lists MPE/iX 6.5 as the build release for Powerhouse. 7.0 and 7.5 releases are also supported, so long as the OS is patched. While some IBM operations have a stellar track record for customer service, Vinnedge said his Cognos experience doesn't match that.
"I have not yet seen a lot of diligent customer service practices, at least on the Cognos side of IBM," he said. "For example, my boss started receiving the 2010 Powerhouse support renewal notice every 15 minutes. It took over a day for my boss to be sure that Kenneth Robinson at IBM knew that there was a problem — and two more days for IBM to fix it."
Vinnedge added that IBM has also changed the Powerhouse license to "Quantity 50 - E06CHLL" without telling Accuride. Even a simple request to explain what the new license meant didn't get a reply. "He has a good guess, based on our other IBM licenses," Vinnedge says, "so he has not really tried to shake that tree."
June 15, 2012
Oracle's legal jousts missing Media's marks
HP and Oracle have been squaring off in court over the future of the database on HP's Unix servers, jousting since the first week of the month. But after a break on Wednesday to attempt to settle the battle out of court, these two companies were back at it after talks crashed. Oracle's got HP's database futures in its hands, and testimony from its executives asserts those hands have crimping sales of HP's Itanium Business Critical Servers.
But that's just not enough to keep the attention of some Itanium owners. One migrator is already heading away from HP's Unix and onto Oracle's Unix. But the death-knell that Oracle wants to spread about the HP-UX platform isn't spooking Greg Barnes.
Barnes has an 3000 background that dates back to MPE III, but his company took its time getting away from 3000s. Media General, which agreed to sell off nearly all of it newspapers to a Berkshire Hathaway subsidiary, was using publishing software from Collier-Jackson until the late 1990s, when the shift started off 3000s. Barnes said the Oracle jabs didn't rattle him a bit in HP Unix management.
"I'm not aware that the snipe-fest had any effect whatsoever," he said, while reporting on the company's in-house migration to the Itanium servers. "Like much of the death knells I’ve read over the years, I have better things to ponder." Among his new interests are Oracle's direct competitor to HP-UX, Solaris. Media General is now phasing out the five HP-UX systems left in its datacenter. It's also focusing on its TV business, post-newspapers.In a shop that runs "an awful lot of Oracle," migrating away from those CJ applications led Media General to Solaris. Peoplesoft HR and FIS, as well as DSI Circulation, either required or recommended Solaris platforms. HP has said for months that Oracle's been attacking the Unix business with success, spreading doubts. But the FUD from Oracle, and its pointing at a sketchy future for the only processors that run HP-UX, was not clouding Media General decisions. Replacing the likes of Collier-Jackson's AIM -- the last of the 3000 apps to go -- determined the OS platform. It was the apps, not how long the chips would be enhanced or supported, that steered every decision.
"At 15 years [of migration], it was very gradual," Barnes said about a shift that started in 1997. "CJ AIM was the last to go, and the new app Mactive also required Solaris."
We originally ran CJ Payroll and AIM. The Circulation and FIS systems were home-grown. They all had their limitations, and management were looking for more features and less programmers.
Payroll was the first to go and that went to PeopleSoft using Oracle. That transition was a major zoo. Next went FIS to the same PeopleSoft/Oracle arena, and that was slightly better. Then AIM to Mactive. Circulation went from the HP 3000 to HP-UX using DSI. We are still migrating it from HP-UX to Solaris.
I was the last man standing to manage MPE: Two 969KS-320s running MPE 6.0. Now I'm the last man standing managing both HP-UX and Solaris, plus Solaris x86 OS.
Barnes reports he was working on HP 3000s just four years after the systems had their true 1974 release. While he jokes that he thinks he's getting entirely too old for this kind of churn, he does have four major operating systems in his resume. "Try managing both Unix and MPE systems, remembering the commands, and not messing up anything if you want a challenge," he says.
Oracle's been meeting challenges of its own since it purchased all of Sun, both software and servers. The latter have seen large sales dips since the 2011 acquisition closed. Selling Unix to new sites has become a trial for HP as well. It's that Oracle database that's got a comfortable spot in places like Media General.
What's slipping away are owned operating profits of more than $2 billion a year off HP's Itanium business that relies on Oracle. A memo from HP's enterprise server chief Dave Donatelli in 2010 said these servers were more profitable than HP's massive PC business. According to an article from All Things D's Arik Hesseldahl, HP's been drawing about 15 percent of its earnings from Itanium business.
HP’s Business Criticial Server business combined with its Technology Services business, which includes the support and services associated with the Integrity line of servers that uses the Intel-made Itanium chip, was at that time larger on a revenue basis than HP’s personal computer business.
And even if HP prevails in its suit, Whitmore isn’t seeing much benefit: “Regardless of the outcome of this particular suit, we expect HP-UX customers to continue fleeing what is increasingly looking like a dead platform — creating a major headwind for HP’s medium-term earnings.” Ouch.
June 06, 2012
Migrations replacing apps, working in-house
3000 migrations are continuing at companies that choose to follow one of two paths. At the Visalia school district in California, operations for its 40 schools will continue to be served by the QSS K-12 applications they've used on 3000s. But IT manager Al Foytek said that Visalia will be following the QSS customers who are moving to the Linux version of the app.
"We will be moving as soon as that's ready, and the front runners are poised for this year," Foytek said. "We won't be in the first wave or two." Schools have a narrow window during the year to make such a transition, typically in the summer between school years. "July 1 would be the ideal time," he said.
He added that Visalia won't move this year and is not likely to make its transition next summer, which would mean its 3000 apps may be migrated to the Linux version of QSS software by 2014.
Customers with packaged applications see advantages to migration beyond just having commodity hardware supplies. Foytek said that moving to an SQL database is also a plus. But QSS has engineered that change for the app, rather than an outside firm selling services to the district. QSS adopted the Postgres SQL database for the Linux version of its 3000 app. Founder Duane Percox has hired key developers from HP's labs such as Jeff Vance and Mark Bixby to do work on the new app's technology.
Foytek said the decision to leave the migration engineering to his packaged app vendor was easy. "A migration of a major system like this is very painful," he said. "QSS provides the lowest cost, by far, solution for schools. Their next competitor is twice as expensive. They manage to hold their prices for software as well as support phenomenally low. If that wasn't true, a lot of us wouldn't still be on the HP 3000. Theirs is an application system that works. Similar systems don't have as many refinements and reporting."
In-house applications — which are best understood by staff rather than service consultants — don't provide the migrations they once did. At the MMFab company in the Los Angeles South Bay area, the maker of fabrics is migrating to a fresh packaged app on another platform. The assistance will be coming from the app vendor, in large measure.
"We aren't converting the old system," said system manager Dave Powell. "There's no business for any of the traditional MPE migration solution providers, so we won't show up on any of their business results. The HP 3000 may keep running for quite a while for history lookups. We aren't migrating historical data like invoices.
"The plan is to move everything off the 3000. Our in-house mail system is separate enough that we could run it without the rest, but I think they are planning to switch to plain-old email — compared to a lot of our custom apps, moving off our mail system should be easy. Everyone has email too, and the in-house stuff is just for some special purposes."
Powell, who's been working on the 3000 30 years, said that programs to send data to the new system regarding inventory and customer files are done and tested. "They just need to run one last time, closer to switchover day." The system will make the leap from traditional 3000 software — only MPEX is running alongside the COBOL that MMFab's developers have customized — to the cloud.
At MMFab, production differs from most manufacturers. "Production to us is placing an order for one of our designs to be made for us, then keeping track of expected completion date, ETA, shortages, quality problems, do-overs, and so on," he said. "It's much different than what I expect most other companies do — no bill-of-materials, for example."
This kind of production — order, invoicing, and sales commissions — all are on the HP 3000. AR, AP and GL live in an old Windows-based package, Powell adds, which is also being dumped. The Windows transition looks easy in relation to the HP 3000 migration already underway. But the nature of such migrations to packaged apps leaves Powell skeptical about a swift cutover.
"The people selling and setting up the new package say things will be ready in by June. I don't believe it," he said. "There's custom programming that hasn't been finished, much less tested, for stuff like designer royalties that their package doesn't handle. User training is in its infancy."
Although the migration at MMFab has eliminated the chance for provider service, the company didn't wear blinders about its method to migrate. In 2011 Powell invited the community's migration vendors to bid the business.
"Consultants with experience with emulation are welcome to contact me to share experiences and fish for future employmment," he said back then in a post to the 3000 newsgroup.
"Whatever we do, it will probably involve more work than I can do myself. Solution providers are welcome to contact me with promises that they can do everything but tie my shoelaces — and especially with promises to enhance their products as needed to do whatever I find out the hard way that they cannot do."
The wisdom from decades of 3000 experience still has value in a migration, contributions which might come from either inside or outside of a shop. Power of Sector7 said the community has always been a pleasure to engage in such large projects.
"We love the old HP 3000 guys," he said. "In today's software world, endless redevelopment is normal. Us old timers instinctively know how to do the job right, and do it right the first time. That is the key to migrating applications — using engineers who just know what to do."
May 14, 2012
Powerhouse drives users toward transition
Fourth generation languages may well be an artifact of a classic time in development, but 4GL code still powers some 3000 applications in enterprises. Powerhouse is the 4GL with the widest installed base, and some of its users are wondering how much time is left on the clock for this advanced development tool.
After its genesis as the Canadian company Quasar, Cognos released and developed this range of tools during the '70s and '80s for HP 3000 reporting, screen design, data dictionary work and applications. At first the Quiz report writer ran standalone on thousands of HP systems, including a bundle as a part of MANMAN's services. But when QDesign, Quick and QTP made their way into companies along with Powerhouse, the whole lineup wrapped itself around commercial apps such as the Amisys/3000 healthcare software -- plus many an in-house 3000 app.
Powerhouse users aren't holding out much hope for improvements to the tool which was purchased by IBM in 2007 along with Cognos. This Advanced Development Tool software didn't drive the IBM acquisition -- the Cognos Business Intelligence tools motivated the purchase. Established Cognos managers retort that ADT continues to produce profits for this business unit. Support contracts for even the smallest of HP 3000s run more than $500 monthly, revenue paid for service now called Vintage Support.
The good news is that Powerhouse for MPE/iX has outlasted Powerhouse for the IBM AS/400, in any vintage. But the language labors under the same yoke that COBOL carries, a profile of a tool built for another time. "The PowerHouse business has to have seen substantial decline for IBM over the years," said Vaughn Smith, a consultant in Canada. "How many more sites can convert to other development environments, reducing IBM's revenue, before they shut down Cognos?"
Smith wrote on a Powerhouse mailing list that "With the exception of Unix and Windows, Powerhouse runs on antiquated hardware." This consultant working with OpenVMS took the official HP view of the 3000, saying the "3000 MPE is done; HP offers help to move these sites to Unix or Windows platforms." (Those 3000 vintage support customers might want to correct his view.)
But even community members with direct 3000 migration exerience see Powerhouse as a waypoint instead of a destination, even when a system built in the '80s would cost millions to replace. Charles Finley of Transformix reported that a high-dollar replacement cost "does not ensure anything" about application longevity.
One prospect hired their web content developer to "completely replace" a working application in six months, because the developers assured them that the 300-program project could be replaced in that amount of time. This was done against the advice of the existing developer and, initially, without consulting her. Four months into the project the web developer asked the programmer for a printout of the database structures. They were TurboIMAGE schemas, so they needed the HP 3000 developer to explain them. The VP running the project who'd hired the web developers suggested that they print out all of the data in the database and have volunteers do the data entry. When the programmer pointed out that there could be lots of errors, she stopped getting invited to the meetings.
I last heard that the system was finally going into testing two years late. What did that cost? This was a non-profit and they did it to save money! Also, as an extra incentive they would have nice web screens instead of those dull terminal screens.
Finley didn't mention the prospect by name, but those details match up with the migration situation in 2010 at the US Cat Fanciers Association.
Costs to carry Powerhouse forward are not a show-stopper for some companies leaving the HP 3000 -- an article in our print edition this month examines such a shift toward Powerhouse on Linux. But the world has changed a lot since the Cognos products were re-engineered in the late '90s to include separate versions for the Web and the Axiant Windows toolset. Much of the product line demands runtime licenses.
One developer who's preparing to make a move to Oracle on Windows and Linux outlined his work, as well as the reasons for doing it. "Once we are fully converted, I expect to start replacing QTP extracts with Oracle stored procedures," said Ken Langendock, "then replace screens with an HTML version that simply gets the data."
I believe, at the end of the day:
1. There are only going to be three databases left: Oracle, MySQL and SQL Server.
2.There are only going to be two OS left: Windows and Linux, because they can be implemented rather inexpensively.
3. There will only be one look and feel for all applications: Web
If Cognos wanted to get back into the running, they would have to follow these assumptions and revamp (combine) all the products into one suite and stop charging for Runtime licenses. They would then have a leg up on all the other tools with their Dictionary, but I don’t see this happening.
May 07, 2012
App replacement may spur emulator evals
The 3000 community continues to examine the Stromasys HPA/3000 emulator from a capital cost perspective. It sounds like a costly investment for a customer who's already working on a limited budget. But there's some evidence from migration prospects that the $50-$100K price tag for the software and Intel hardware may be a price that can bridge the timeline to app replacement.
Users and managers in the Powerhouse community have been studying the future of remaining with HP 3000s, some 16 months into a period where Hewlett-Packard stopped providing support for the OS and hardware it created. While the debate included one "get off, anyway" opinion -- a consultant said Powerhouse users should "put a bullet in this OS and IMAGE database and move on" -- another view is that the emulator might be a stopgap for replacements.
Anne Quirke of the Dublin, Ireland consultancy Uturn Ltd. said that one client prefers to replace an app now on the 3000, instead of migrating it. Replacement is a different set of costs and efforts than lift and shifting business apps. But it still might spur some attention during plans to sustain computing resources.
Reducing these risks around the PA-RISC iron is a reason cited for the move away from the 3000 platform and MPE/iX. Quirke said the company considers the Stromasys product "one option we are considering looking into." That's language which suggests months may pass before that 3000 could be replaced with an emulator. But hardware issues rise up during the examinations of 3000 futures.
Migration is not an option for a long, long list of reasons; replacement to a new application is preferred. The time-line to these new applications is not directly in our control, so in the meantime we are looking at options to reduce the risk associated to the hardware.
"We are have been in situations where we have to wait and pray that replacement parts can be found," Quirke said, "or we rob parts from test boxes. As hardware availability and support is a key factor which brings the migration discussion to the fore, virtualization, on paper, offers some solution."
April 27, 2012
3000 system census surprises in UK
At a recent 3000 webinar among CAMUS user group members, the Talk Soup Q&A brushed across the 2011 HP3000 Reunion. While the talk examined activity of 2012, one attendee on the conference call could be heard saying, "Not another reunion!" It's a tiresome but expected response to the scope of the 3000 population.
On one side stand the users and managers who employ an HP 3000 in everyday production. They're grateful for any relevant information to keep 3000s running well and updated as much as possible. These community members don't often ask how many systems are still running. For some, another Reunion would be a chance to attend an event they couldn't enjoy because of a 2011 conflict.
Other HP 3000 managers want to view the community as a seriously shrunken village. They've made the choice to migrate, or they can't find work any longer that taps their MPE and 3000 skills. Perhaps they do business in the community and haven't had new revenue in a long while. Other opportunities call, so they're eager to reinforce their choice to move away.
However, we sometimes encounter census trail-posts that lead away from the "too small to be relevant" viewpoints. In the UK one prominent community member had a trail blaze that opened their eyes about who might still remain in the homesteading populace.Like most of these reports, it came by way of a third party. One vendor said he saw a support provider's list of 3000 sites and spotted suprising totals.
"There were twice the number of customers on that single vendor's list as I suspected were in the entirety of the UK," he said.
One thing that might well stall migrations this year -- and sustain that populace -- is the emergence of the HPA/3000 emulator product. This virtualization engine won't even have to show much success at this point in the 3000's life. With a solution other than migration on the horizon, the lean-budgeted 3000 users will have something to use as a risk-aversion strategy. A company with concerns over hardware availability or costs will believe that hosting MPE on commodity hardware resolves those problems.
Whether that's reasonable or not remains to be proven. In the short term, the hardware suppliers to the community will survive, because the costs of shrink-wrapped replacement components will remain well below the fee to install HPA/3000. Even a cloud-based deployment will cost more than fresher hardware with HP's badge. Given enough time, the $25,000 entry to that commodity solution may seem a better long-term strategy.
It's a common belief that a 3000 emulator arrived too late to make a difference in the market. But learning that twice as many customers remain online as expected changes that formula -- especially for each customer who's remaining a homesteader.
April 25, 2012
Continuing support key to homesteading
In a webinar last week the makers of the HPA/3000 Charon virtualizing engine (read: emulator) took questions from attendees about licensing. Not the license of MPE/iX (already in place from HP) or licensing their product with customers (something they'd love to do once a customer commits. Soon, we were told.)
The licensing issue in play is how to get a software vendor to embrace use of their product on HPA/3000. For some companies this is an automatic. They generally don't charge for upgrades and haven't created anything that needs special handling inside MPE/iX. Terry Floyd of the Support Group sells software that his company has crafted. His customer, Ed Stein of Magicaire, is on the short list for early adoption of HPA/3000.
"I don’t write any tricky stuff," Floyd said. "We don’t have anything that needs testing. If Ed could get a box with Charon running, our test would be a full month-end close (dozens of jobs) and an MRP run. I think he’ll do a very thorough job – that’s his nature."
Some vendors, especially app suppliers, might have a different approach. The key to getting software from HP iron onto the emulator may well be keeping up support. 3000 software support contracts can be left behind while trimming budgets. This can present a problem that can be fixed by restarting support -- which is a good idea anyway, if the 3000 is mission-critical.The flow of support money is tricky. For some 3000s this expense will need to be justified. Floyd sketches out the issues, both from the customer's point of view as well as software vendors who still support 3000s.
Nobody who's on support will have any problems with a vendor. So the real crux is: are you on support with each vendor? If not -- in Ed’s case we can use Cognos for an example -- they probably are not going to be very concerned that they intend to charge you to get back on support. Perhaps a considerable amount. As someone who lives on support income, I guess I can feel their pain, But we have never charged anyone extra for back support if they left us, then came back later.
What is somewhat comparable in our arena is charging higher rates to do any work for MANMAN users not on our support contract. That is our policy and we haven’t broken it yet. I guess some would think that is hard-nosed, but if Ed has been off support for, let’s say, 10 years on Cognos' Quiz and now wants to go back on support with them, they might begin the negotiation with a charge of the full 10 years back support, Just to get his attention. I can see -- for some HP 3000 users not on support with some vendors -- why this is going to end up being an “every man for himself” negotiation process.
April 23, 2012
Federal program helps 3000 IT pro re-train
HP 3000 IT pros have a challenge to overcome in their careers: how to add modern skills to the classic tooset they learned managing 3000s. Those between jobs must handle the costs to train, too. Craig Proctor has been spending time to learn the likes of C#, Java and Visual Studio. After a year of study, he hasn't been spending his own money.
"I took a dozen different classes," Proctor said. "The Trade Act paid for it all. It's possible to take one class at TLG Learning, or work with them to design a series of classes."
Proctor worked with a 3000 for more than 20 years at Boeing, as a Configuration Management Analyst and Business Systems Programmer Analyst. He left Boeing in 2010 and began a period he calls Updating IT Skills in his resume at LinkedIn. TLG, based in Seattle, gave him training that he will blend with the business analysis that's so common in 3000 careers. He understands that by drawing on his recent education he'd accept at an entry level IT position. "You get the merger of an experienced analyst, using new tools," he said of his proposal to any new employer."
Last year an extension of the Trade Act was signed into US law by President Obama in one of the few bills that escaped the partisan logjam. A federal website describes it as a way for foreign-trade-affected workers to "obtain the skills, resources, and support they need to become re-employed." $975 billion in federal funds have been sent to states like Proctor's in Washington, adminstered by each state. Furloughed workers file a petition for training, job search and relocation allowances. These pros have an average age of 46, which is the younger side of the HP 3000 workforce.Proctor didn't believe that his 3000 experience helped in gaining more modern IT skills -- except for his years as an analyst.
I wouldn't say that the HP 3000 skills helped, but the analytical/programmer skills did. All 22.5 years at Boeing were on the HP 3000 (Fortran) and I had a couple of years on it before. as well as Burroughs (now Unisys) using COBOL. The hardest class for me was C#; COBOL and Fortran were so similar, but C# was nothing like that. The other classes were interesting and fun for me -- challenging, but still fun.
Like anybody well-versed in system management and coding under MPE, he'd like to land a job in a business using a 3000. "With so much HP 3000 experience under my belt, I'd feel a lot more comfortable and ready to dive in with another HP 3000 shop," he said. "I also have all the soft skills -- investigative, detail oriented -- that I need."
Learning what Proctor called "21st century technology" can help 3000 veterans who've seen their positions eliminated. There's a LinkedIn Group devoted to HP 3000 Jobs with more resources and discussion. It's a subgroup of Bill and Dave's Excellent Machine, devoted to the HP experience. Like the HP 3000 Community Group, (now 475 members strong) you request membership -- but a 3000 pro sees nearly-automatic acceptance in these groups.
April 16, 2012
Migration racks up list of emulated tasks
Some HP 3000s which remain in service are using many MPE nuances to get their jobs accomplished. Each of these tasks needs to be emulated in a migration away from the server. Even as companies embark on migrations to reduce risks, the list of tasks that they hope to replicate from their in-house apps can be surprising.
Such is the case at MM Fab, a fabric manufacturer in LA's South Bay Area. The 3000 shop is now taking its first year of steps off the system, developed and managed by Dave Powell. He shared a list of the things that an emulator must do if it were to succeed at replacing HP's 3000 hardware at his shop. The list also serves as a extensive catalog of the capabilites required of any new operating environment.
"We are thinking about migrating," Powell shared, months before the decision was made. "Which means we have to think about the choice between buying a package vs some form of emulation. Which means I could use some assurance that the [3000 hardware] emulation tools out there would actually work for us."
I can't afford to take this for granted because our system uses some rare features and does unusual things. Lots of them. Example: we do lots of tricky escape-code screen handling (mostly for point-and-shoot, drill down inquiries) that breaks some terminal emulators. Reflection 10.0 works, as does Minisoft WS92 v5.4 and actual terminals from 262x on, but last I checked, Minisoft Secure92 fails big-time. Not trying to make Minisoft look bad, but I need to make the point that software that works elsewhere may not work for us.
"We never cared about portabililty," Powell said, "because we never had any intention of moving to any other platform." From such situations are customers made for the Stromasys virtualization engine. If you're uncertain of whether you're using any MPE nuances in your application, it's a good strategy to get an evaluation of what's in production use today. Even if you're not migrating.Powell said he doesn't think terminal emulation will be a big migration issue. In an emulation, "I think we could just keep on using the two products that work -- I just need to emphasize that we are off the beaten track, feature-wise."
Since there won't be as much room for all the details of MM Fab's custom-code tricks in our printed edition, we thought we'd put them on display here. This list might be useful to let you see if any of this is working inside your in-house apps. For the record, Stromasys says that anything that's working on MPE today will work in its emulator. The only exceptions they've found were HP's internals diagnotics, like SHOWCLOCKS.
A new platform/replacement app would have to embrace the top-level abilities in Powell's custom-code list. It's the kind of situation that makes some 3000 customers a poor fit for a migration, because these nuances were built over more than 20 years of IT budgets. A migration or replacement would address these all at once -- a cost structure that many 3000 shops cannot endure today.
Powell's MPE magic:
Job queues with separate job limits.
Smart :pause command (wait up to 'x' seconds for that job to log off).
MPE functions like finfo and jinfo.
User functions. Some of them are extra date / calendar routines beyond the built-in ones, like "how many days till end-of-month?" or "how many work-days in the next 'n' days?" and "how many months old is this file?"
MPE variables. User variables plus system variables like hpdatetime, hpaccount, hpfile, hpcpusecs, hpjobcount, hpstreamedby.
Message files / circular files / temp files, including temp message files and temp circular files.
Lots of command files, with tricks like with multiple entry points, input or output or both redirected to files, etc. Command files that use :echo to build a job (in a temp file) which they then stream. (I always wanted a way to have UDCs/command-files run offline, or feed parms into a job like UDCs do, so I finally rolled my own).
Jobs that use :echo or :print to build command-file subroutines (also in temp files), which they can then call lots of times with different parms, like running the same program over and over with one cmd-file parm becoming the info that is passed to the program to tell it what to report, another parm becoming part of the file name where it stores the report output, and another parm telling it who to email the report to.
Lots of do-it-yourself logging, with overglorified :echo to circular files, so I don't need to worry about the logs getting too big.
VPlus, with heavy use of vchangefield in newer apps, and family-of-forms in older ones, both to dynamically make some fields inputable and others display-only, changing the display enhancements so users can see which is which.
Creative escape codes in vplus apps to do things that VPlus didn't do as nicely as we wanted, mostly setting function-key labels and screen-printing.
Lots of escape sequences in non-VPlus terminal IO, mostly in character mode.
Extra terminal control features like turning echo on and off, time-out reads, etc. (Hint: escape codes that cause the terminal to send data back to the computer may work most of the time, but don't get solid unless echo is off. Even so, if something goes wrong you don't want the computer to wait forever for an answer).
Lots of env-files for both lasers and old impact printers, mostly changing orientation, print-size and lines per inch so the same report can print on either type of printer. Some reports have a run-time way to tell them how many lines per page, so by coordinating that with env-files I can have a report that normally does 132-column 60 lines do really-small-print portrait mode 124 lines per page on a laser. Also some tray-selection in env-files.
Do-it-yourself fancy laser-printed invoices with legalese in very-tiny print, company name in big print, etc. No special forms package here, just me spending quality time with the PCL documentation.
Converting simple report output to PC-readable format. That's a one-liner on our 3000 with my HP2RTF command file. The new system doesn't have to use RTF, but it does have support a common PC-readable format, has to preserve/translate HP-style line-spacing and page-breaks, and has to support changing print-size and line spacing so the PC file will look normal on screen and printed page. And it has to be easy to invoke in batch.
Email reports. This is also a one-liner here, thanks to a set of command-files I have wrapped around a nifty mail program originally from Telamon. The command files provide logging, improved error-checking, distribution lists, and even automatic retries at gradually-increasing intervals if there is an internet connection problem. I would like to keep that functionality. If possible I'd like to keep the outer layers of my command files, wrapped around whatever mail-sending pgm exists on a new system.
Mass file rename/delete/print/email, with ability to select by date, file age, file size, etc. Some use MPEX, others use my own routines (listf into a file, read it back, maybe call finfo).
IMAGE b-tree dbfinds.
COBOL macros. Intrinsics like command. Any and every HP extension that ever seemed helpful over the last 30 years.
April 11, 2012
Changing IP Addresses for HP 3000s
I need to change the IP address of our HP 3000 in the near future, and it's been over 10 years since I've done anything like this. Here's what I think needs to be done:
Put in the network interface, (LAN1), then press Config Network
Enter the new IP address
Tracy Johnson replies:
I would go with Unguided Config. Guided may change things (besides the IP address) to defaults that may have modified over the last 10 years.
Craig Lalley adds:
Depending on the old IP address and the new IP address, you may want to also change the subnet, and the gateway. The gateway can be accessed by hitting F4 for Internet. The gateway is found at the path NETXPORT.NI.LAN.INTERNET
If you are making the change because of a new switch/router, make sure the network guys configure the port for the HP 3000 correctly. In other words, if you have a 100MB card, make sure it is set to 100MB/full duplex and do the same on the HP 3000, and turn off auto negotiate.
Independent IT consultant Al Nizzardini adds that creating a new System Load Tape is an important part of the process. Gilles Schipper of indie support company GSA also explained a key step.
After making the change via NMMGR and validating both netexport and DTS, you need to:
to actually effect the change
April 10, 2012
Manufacturers pull HP off support lines
CAMUS director Michael Anderson, an IT consultant in the Bay Area and a leader of that MRP/ERP users group, was an IT projects manager and applications manager at manufacturers Tencor and ThermaWave, both using HP 3000s. Hewlett-Packard is off the radar at most of these manufacturing sites.
“As far as HP support for the HP 3000, I dropped mine a couple months after they announced the end-of-life,” Anderson said of the period in 2002 where he became Enterprise Wide Applications Manager at ThermaWave. “What are you spending money for at that point? Long term there was not going to be any meaningful development for MANMAN, so there would also be no demand from the application for new features in MPE or IMAGE. My employer was going through tough times and really needed the $58,000 in savings.
“As long as the old hardware continued to work and you had a good boot tape, what could HP provide that third parties didn’t already provide better for less? The damage is done and most of the HP support customers are gone. Maybe if they had announced there would be some support in the afterlife there would be more users holding on.”
User groups, which have some of the most seasoned managers in the community, offer a better application and system resource. “On the other hand, for the companies that still use MANMAN on HP 3000s, CAMUS is still here to provide a supportive environment and forum for knowledge exchange,” Anderson said. “But it’s getting pretty quiet.”
While HP’s not making much noise on these soft feints into a market that it’s abandoned, there’s no doubting the attempts will continue, however unsuccessful. MB Foster’s Birket Foster predicted back in 2009 that HP would become a non-entity in the support field by now.“Does the market miss the final level of HP’s 3000 support? No, these customers are already working with independent companies,” Foster said. “I’m sure that the only thing that annoys those [independent providers] is that HP keeps taking money for support. The long support tail of HP has already moved resources away from the 3000.”
Off the books, however, the HP methods and pricing are still being applied in some places. “They’re still doing support for some customers under nondisclosure because they don’t want the unwashed masses to know,” Foster said. “They’re willing to ensure there’s a body providing support to those customers. It’s something they’ve managed, to keep employees on for an extra period of time to cover some of the support needs for some larger, more strategic customers. But HP is also working hard to ensure those customers have a plan to move off.”
“They’re definitely there on the hardware side of 3000 support,” Foster said. “They’re offering support, but for certain devices. The list is growing smaller. When you get an HP renewal, it now says, ‘Except for these devices,’ with another set of devices falling off the list. For operating system support, I’m sure you can get it if you’re paying enough money.”
“Some of the people who used to work on the 3000 are still working inside HP. They’re very experienced and support certain customers. But I would say HP’s definitely reduced the number of personnel skilled in HP 3000 support.”
The departure of HP from the field follows a pattern of receding that started long before the vendor closed up its 3000 labs. At the MANMAN and 3000 support provider The Support Group, “We felt like we were supporting legacy products already in 2001,” says founder Terry Floyd. “Most of our MANMAN customers were off of applications software support anyway, so it didn’t change our plans much.”
April 09, 2012
CAMUS webinar includes emulator update
The CAMUS ERP/MRP users group is hosting an online meeting in about a week, on April 17 starting at 11:30 EDT. CAMUS board member Michael Anderson is taking registrations for that Tuesday's call-in and web briefing, one which includes an update from the makers of the Charon HPA/3000 emulator.
Stromasys has added an HP 3000 business manager, Paul Taffel, who will brief attendees on this HP 3000 emulator. Taffel's got airtime on the docket through 1 PM. A demonstration is promised. The meeting is open to anyone who registers with Anderson, by sending him an email. He'll reply with login and call-in details.
In the hour following the Stromasys briefing, users who are managing VMS sites will share information in a Talk Soup about the track record of the Charon technology in the DEC world. The first ERP-MRP production work for the emulator took place in the Alpha and VAX community. Some CAMUS members have already shared high praise for the software's ability to mimic HP hardware (on VMS systems) using Intel PC systems. What's changed since those Charon versions is the hosting environment. It's now Linux instead of Windows.
Anderson says this spring's premiere of the HPA/3000 offer may not fit the users of older, smaller 3000s. The first release of HPA/3000 is only matching A-Class 400MhZ horsepower. Stromasys has proven lab resources to boost that. The rollout schedule promises an N-Class-powered, multiple processor version by sometime after July 1, but sold at a price above $50,000.Anderson has said that based on the first set of prices, he doubts there's enough of an offer yet from Stromasys to spark growth of MPE application use during 2012.
"I don't see the emulator swelling the ranks of MPE or application users," Anderson said. "I think they are mistaking the ardor of the enthusiast as market demand."
"I think there's a lot of interest in the emulator for contractors who can finally put a development system into their home office," he added. "I don't think we're to the point yet when you can't get parts to keep old hardware running — just rob them from an equivalent HP 9000."
There's also a general discussion of CAMUS activities set for the meeting. Setup for the webinar is scheduled to start at 11:15, to ensure smooth connections.
March 27, 2012
Protecting HP 3000s Using Linux
While HP 3000 sites deploy Linux servers this year, some of them are using the environment as a buffer for 3000s which need to be in range of the Internet. James Byrne, who's hosting the hp3000links.com website as well as managing IT project for Harte & Lyne, outlined his setup to use Linux for 3000 protection.
Byrne has his HP 3000s and the internet buffered by a dual-homed Linux box in front of the HP 3000, using that to provide firewall, SSH, and proxy services. He describes his setup a fairly primitive (where GW/FW=gateway/firewall):
Internet-> GW/FW <-> Eth0:Linux:Eth1 <-> HP 3000
The network connection to the gateway/firewall provides our public routable access. The link between the Linux front-end host and the HP 3000 is a x-over cable using a 192.168.0.0 block address. Direct network connections to the HP 3000 NIC are physically impossible. This ensures physical network security over the non-encrypted portion of the network (for SSH access).We use a CentOS-5 based host running IPTables, Squid, OpenSSH, VSftpd, and Denyhosts as the front-end to the HP 3000. IPTables is configured to log and drop for 7 days all addresses performing obvious port scans. IPTables similarly counts, logs and blocks IP having excessive failed connection attempts on visible ports.
There are a wide assortment of Linux-based firewall appliance distributions which may simplify set up somewhat for novice users. Alternatively, one can simply use a mainstream Linux distribution, or a derivative like RHEL/CentOS or Debian/Ubuntu, and add and configure the packages desired.
Denyhosts scans the logs for other issues and really does not add much to our setup. However, Denyhosts can be used to do itself everything I have chosen to do in IPTables. Therefore, one may concentrate on learning the configuration of just Denyhosts and leave IPTables configuration to the minimum necessary to allow access.
The proxy server handles FTP but we do not allow FTP access to the HP 3000 at all -- so I could not tell you if we have that set up correctly or not. We have it there in case the need ever arises.
The intellectual load of dealing with these things is non trivial. However, the price of freedom is eternal vigilance. Once the front-end is setup ,we run logwatch to send daily reports on connections and consider whether further configuration changes are necessary.
March 06, 2012
Assisting Off the 3000, En Route to Linux
A worldwide travel and healthcare insurer is making the move off their HP 3000 starting this year. While that's not remarkable, the destination is notable. Europ Assistance is starting the work to replace its MPE host with a Linux system, right down to considering a Powerhouse license re-purchase.
Adrian Hudson is part of the IT team at Europ, a firm which sells insurance for travels as well as supplemental healthcare. Since these policies are purchased one-off, as the UK-based firm might say, customers pay for them with credit cards. That's the spark to replace the HP 3000 with Linux, Hudson says.
"As Europ Assistance is involved in the Payment Card Industry, one of the key drivers for the migration away from the 3000 is regulatory compliance," he reported. The PCI regulations have been a challenge for some companies to master using the 3000. Last year Hudson was researching a way to permit the HP 3000 to process payment card information using Secure File Transfer Protocol. SFTP was not entirely supported by HP prior to the Hewlett-Packard lab closing in 2008. Hudson was diligently working on a way to involve the 3000 in these data transfers. The alternative, to use intermediate SFTP support on non-MPE servers, turned out to be the solution.
"We ended up piggy-backing files through a Windows server with SFTP installed," he said, "and then FTPing them to and from the 3000." Now the operations once handled by that 3000 are heading to a Linux server. Hudson is investigating the cost of keeping Powerhouse in place on the application. It's one of the simpler ways to migrate code to an alternative platform.One of the first steps, with the most exciting outcome, is discovering what the charge to moving to Linux will be on the Cognos price list. In the years since Europ Assistance first bought Powerhouse, Cognos has become a part of IBM. The 4GL -- it's called an Advanced Development Tool -- is licensed differently for Linux developers than on the MPE/iX systems.
We received a quote from IBM and it is for a license per ‘named’ developer. I understand the licensing structure is now only for a development license and there is then no need for a ‘Production’ license.
In the past, one would buy a Full License for the Development machine and a Runtime (maybe with reporting) for the Production box, and I believe there was no limitation on the number of users.
The Powerhouse product manager Bob Deskin explains that the definition of user under Linux Powerhouse has shifted only slightly from the classic 3000 terms.
The license model has not changed very much. What you're probably thinking of is the HP 3000-MPE/iX platform, where we would typically license by machine size. Therefore, it was unlimited users for however many you could run on that machine. With other platforms, the approach was to use the number of users. In most cases this became the number of sessions, rather than unique named users. In other words, if a user opened two terminal emulator windows on a PC, that would count as two sessions and two users. The only exception was PowerHouse for Windows, where it was assumed that there was only one user.
Under IBM, instead of sessions, it is truly a concurrent user. And further, they specify named user as in unique user. They do not expect you to name all the users. So under IBM, the above scenario of a single user opening two sessions would only count as one concurrent named user.
There is still a distinction between development and runtime, but it depends on the platform and use. If someone purchases a single development license on Windows, there is no need for anything else. It's a single-user machine. But if you buy a single development license on a Linux server, you require runtime licenses for your users.
License structuring for other development tools on Linux may not require runtime purchases as a matter of course. But it's interesting to note that Powerhouse Linux demands this extra cost, while Powerhouse Windows doesn't. Many migrating HP 3000 sites have chosen Windows as their alternative platform. However, of late many others are looking at Linux -- with its improved and still-enhancing enterprise features -- as their best alternate to MPE/iX reliability. IBM/Cognos might be choosing its license terms in response to the enterprise's migration to Linux. Managers routinely point at Linux's affordability as important to their choice, however.
The runtime licenses you didn't need on an HP 3000 are required for Powerhouse Linux. The overall cost is likely to be less. MPE/iX licenses for Powerhouse were legendary for their cost -- support alone can easily be five figures a year -- and inflexibility during upgrades. Current customers like Europ Assistance, with services and servers on five continents, may be considering how many runtime purchases they can afford to purchase for what was supposed to be a more affordable platform.
Hans-Ole Kaae, an IT consultant, also wants to understand these costs to migrate MPE/iX-based Powerhouse to Linux. "If you have, say, a new or a current customer, heading for Linux or Unix, is this all it takes these days: X developer licenses and Y run-time licenses -- plus, of course, X + Y data access licenses?" Deskin says that's about it.
You may need an extra runtime if you're running batch jobs as a separate user. And data access is per source. So if you're using C-ISAM and Oracle, you would need two data access per user.
Also note that IBM does not distinguish between platform. If you're on HP-UX and move to Linux, you can move the existing concurrent user licenses over as long as you don't exceed the overall entitlement.
Deskin said nothing on the Powerhouse-L mailing list about moving MPE/iX licenses over, because the HP 3000 Powerhouse was licensed by system, not by user. There may be a need for custom quoting to determine how much Powerhouse on Linux will cost. There's support to be paid on every extra license, after all.
February 27, 2012
3000 support demands spare inventory
Independent service providers have signed up most of the 3000 homesteaders by now, according to Pivital Solutions' Steve Suraci. The CEO still runs across the occassional shop served by HP out of habit. A big share of the available service contracts have already been passed to independent companies, however, according to an article in our in-the-mail February NewsWire print issue.
But using an independent firm for support is a smart deal only if the provider has ample spare parts allocated to your site, Suraci said. A system administrator who manages the Series 969 at Hostess Brands (how's that for a large homesteading company -- Twinkies anyone?) needed an HP A5418A fiber router (at left) to replace a blown device. The indie support company serving Hostess didn't have one, so Joe Barnett went looking on the 3000-L mailing list himself. He needed to maintain connectivity to his VA7410 array, or face rebuilding the array from backup tapes.
Solutions and suggestions trickled in -- including the purchase of one 5814A for sale on eBay "that might not rewritable," because it wasn't the MPE -003 model. What's more, that vanilla unit ships on 4-14 days delivery time, according to the eBay listing. Suraci, whose company specializes in 3000s, pointed at a weak Service Level Agreement (SLA) as a bigger problem than just not being able to get a replacement HP router.
The easy questions to answer for a client are "Can you supply me support 24x7?" or "What references will you give me from your customers?" Harder questions are "Where do you get your answers from for MPE questions?" Or even, "Do you have support experts in the 3000 who can be at my site in less than a day?"
How many HP 3000 shops are relying on support providers that are incompetent and/or inept? The provider was willing to take this company's money, without even being able to provide reasonable assurance that they had replacement parts in a depot somewhere in the event of failure. There are still reputable support providers out there. Your provider should not be afraid to answer tough questions about their ability to deliver on an SLA.
But Suraci was posing one of the harder questions. "Here are my hardware devices: do you have spares in stock you're setting aside for my account?" Hardware doesn't break down much in the 3000 world. But a fiber router is not a 3000-specific HP part. Hewlett-Packard got out of the support business for 3000s for lots of reasons, but one constant reason was that 3000-related spare parts got scarce in the HP supply chain.
The economy has recovered a bit, Suraci said, so he suggested now's the time to ask these hard questions. "It might be time for everyone to review their support provider, and maybe look a little deeper than what they charge for service," he said. "In many cases, you get what you pay for. Response time, parts availability, and legitimate HPSUSAN updates all need to be addressed in advance of signing on the dotted line. It's one thing to be budget conscious, and a whole other to be blinded by it."
Even when a last-minute email could solve a parts problem -- and it looked like Barnett might have gotten lucky on locating a spare router -- that's not a reliable support plan. One suggestion was a Crossroads SA-40 switch, but Craig Lalley notes that you can't boot a 3000 via the Crossroads device. He had to hook up a Mod 20 storage unit for boot-ups only.
Jack Connor, who does 3000 support work for Abtech, seconded Suraci's advice. "I couldn't agree more. Costing out the spares and having them available should be part of the contract."
February 21, 2012
Respect MPE spooler, even as you replace it
Migration transitions have an unexpected byproduct: They make managers appreciate the goodness that HP bundled into MPE/iX and the 3000. The included spooler is a great example of functionality which has a extra cost to replace in a new environment. No, not even Unix can supply the same abilities -- and that's the word from one of the HP community's leading Unix gurus.
Bill Hassell spread the word about HP-UX treasures for years from his own consultancy. Now he's working for SourceDirect as a Senior Sysadmin expert and posting to the LinkedIn HP-UX group. A migration project just finishing up drew Hassell's notice, when the project's manager noted Unix tools weren't performing at enterprise levels. Hassell said HP-UX doesn't filter many print jobs.
MPE has an enterprise level print spooler, while HP-UX has very primitive printing subsystem. hpnp (HP Network Printing) is nothing but a network card (JetDirect) configuration program. The ability to control print queues is very basic, and there is almost nothing to monitor or log print activities similar to MPE. HP-UX does not have any print job filters except for some basic PCL escape sequences such as changing the ASCII character size.
While a migrating shop might now be appreciating the MPE spooler more, some of them need a solution to replicate the 3000's built-in level of printing control. One answer to the problem might lie in using a separate Linux server to spool, because Linux supports the classic Unix CUPS print software much better than HP-UX.The above was Glen Kilpatrick's idea. He's a Senior Response Center Engineer at Hewlett-Packard. Like a good support resource, Kilpatrick was a realist in solving the "where's the Unix spooler?" problem.
The "native" HP-UX scheduler / spooler doesn't use (or work like) CUPS, so if you implement such then you'll definitely have an unsupported solution (by HP anyway). Perhaps you'd be better off doing "remote printing" (look for that choice in the HP-UX System Administration Manager) to a Linux box that can run CUPS.
This advice shovels in a whole new environment to address an HP-UX weakness, however. So there's another set of solutions available from independent resources -- third-party spooling software. These extra-cost products accomodate things like default font differences between print devices, control panels, orientation and more. Michael Anderson, the consultant just finishing up a 3000 to Unix migration, pointed out these problems that rose up during the migration.
My client hired a Unix guru (very experienced, someone I have lots of respect for) to set this up a year or more ago. They recreated all the old MPE printer LDEVs and CLASS names in CUPS, and decided on the "raw" print format so the application can send whatever binary commands to the printers. Now they have some complaints about the output not being consistent. My response was, "Absolutely! There were certain functions that the MPE spooler did for you at the device class/LDEV level, and you don't have that with CUPS on HP-UX."
Anderson has faith that learning more about CUPS will uncover a solution. "One plus for CUPS, it does make the applications more portable," he added.
There's one set of tasks can solve the problem without buying a commercial spooler for Unix, but you'll need experience with adding PCL codes and control of page layouts. Hassell explains:
Yes, [on HP-UX] it's the old, "Why doesn't Printer 2 print like Printer 3?" problem. So unlike the Mighty MPE system, where there is an interface to control prepends and postpends, in HP-UX you'll be editing the model.orig directory where each printer's script is located. It just ASMOS (A Simple Matter of Scripting). The good news is that you already have experience adding these PCL codes and you understand what it takes to control logical page layouts. The model.orig directory is located in /etc/lp/interface/model.orig
What Anderson needs to accomplish in his migration is the setup of multiple config environments for each printer, all to make "an HP-UX spooler send printer init/reset instructions to the printer, before and after the print job. In other words: one or more printer names, each configured differently, yet all point to the same device."
You won't get that for HP-UX without scripting, the experts are saying, or an external spooling server under Linux, or a third party indie spooler product. If you'd like to look over the discussion in real time and add questions, it's on the LinkedIn HP-UX group's webpage. The third party software list for Unix is long. ROC Software moved into this field more than six years ago, along with its support of Maestro job scheduling for the HP 3000. ROC's products for Unix are Rhapsody and EasySpooler, for multiple-server and single-server environments, respectively. Another spooler software vendor with 3000 experience is Holland House, which sells its Unispool product for environments including Unix.
3000 managers who want third party expertise to support a vast array of print devices are well served to look at ESPUL and PrintPath spooling software from veteran 3000 developer Rich Corn at RAC Consulting. Corn's the best at controlling spoolfiles for 3000s, and he takes networked printing to a new level with PrintPath. Plenty of 3000 sites never needed to know all that his work could do, however -- because that MPE spooler looks plenty robust compared to what's inside the Unix toolbox.
February 13, 2012
Developers, users manufacture 3000 chat
A lively discussion is in play at the HP 3000 Community of LinkedIn, where users, developers and managers are examining issues around migrating away from an MPE application of serious size and age. Or the need to do so.
Once Randy Thon mentioned he's using MM/3000 to manage maintenance services at Cessna Aircraft -- adding that the company's looking at options to leave the 3000 -- others in the 425-member community supplied advice and counsel.
The options suggested to Thon go beyond using the new Stromasys emulator. He's pleased with the way his app is working on the 3000 for Cessna. The hardware is the burr under the aircraft maker's saddle. The migration of an app like MM/3000 is a project that taxed every aspect of the software's owner, a crew laden with ex-HP engineers.
"The eXegeSys team spent years trying to migrate MM/3000 to Unix and ultimately gave up," said Jeffrey Lyon, "and sold the intellectual property. 11.7 million lines of COBOL, SPL, and Pascal is a big beast to move."Another community member said that 11 million lines of code isn't that large, really. "The 11.7 million is not that big," said Brian Stephens. "I did a migration at Speedware; think it was about 4 million lines of COBOL and 300,000 Pascal and SPL in about a year. Our team was 14 members and we started not knowing the app. A bigger team, knowing the app, could get the MM3000 migration done in under two years."
There's also the issue of whether you would get what you were really pursuing, once you'd complete a migration. These are different issues for a software vendor than a user of its products. Have a look at the chat and chime in with your own experience about migration strategy.
February 09, 2012
Third Party Futures Revisited, Maintained
Early this morning I went on a search for modules of HP's Maintenance Management/3000 software, known as MM/3000. A new member of the LinkedIn HP 3000 Community posted his user profile on that group (425 members and counting), and Randy Thon identified his shop as an MM/MNT user. The software that's running at his HP 3000 site was first installed in 1988. Thon explained that the program suite is still functional and efficient today.
The HP 3000 is still the core of our application. We're running on a Series 969-420 and rebooted two months ago -- we last rebooted five years ago. So far the application has been very robust, averaging production application changes weekly, allowing us to change at the speed of thought to accomodate changes in the manufacturing workplace and reductions in workforce. One of the main reasons we are still on this application and platform is that it is cost effective, solid and all development and management of the system is within the Maintenance Department.
That's the maintenance department of the Cessna Aircraft Company, the world's largest manufacturer (by aircraft sold) of general aviation airplanes. Not exactly a small enterprise, and there's clearly no software problem in Cessna's maintenance group. (Thon, by the way, is looking for fellow users of MM/3000. You can link in to him via the HP 3000 Community.)
The ease of integration which lets Cessna "change at the speed of thought" is enhanced by a third-party piece of software that improves MM/3000. Products like the eXegeSys eXegete client, a front end for the MM/3000 software, have made using 3000s to drive a big company a safe long-term investment. It's been that way for more than 30 years in your market, but there was a time when any software sold outside of HP was a budding enterprise. I located a link to illuminate this pedigree at the Adager website, where long-term 3000 resources have always had a generous harbor.
On the Adager site you can read "The Future of Third-Party Vendors In the HP3000 User Market." The paper written by Eugene Volokh of VEsoft at the end of 1983 does some in-house forecasting. Third parties are going to do well in the world of 3000 owners, Eugene figured, because the system vendor would always be missing out on improvements, innovation, or competitive pricing on software. This might seem like a no-duh theory now. But in the world of 1983, independent providers of computer solutions were anything but a slam-dunk in the world of enterprise IT.
Volokh, Adager and Robelle are among the group of software solution "Improvers" that Eugene cited in his historic paper. In essence, after 3-4 years of success from these companies the case was pretty well proven that a solid product like MPEX, Adager, Qedit or Suprtool was going to win a lot of business away from the systems makers.
But the point that you might overlook in the paper is that these three companies continue to make long-term investments in 3000s possible and profitable, even after three decades. Eugene was just taking note of a software trend that remains true today: innovation from outside the system creator builds a lifelong community of support.
In a recent talk with Birket Foster, whose MB Foster Associates celebrates 35 years of continued business this year, he reminded me of where the community turned for new ideas in the early 1980s. The third-party vendors such as Foster, Adager, VEsoft and Robelle turned out papers, published books and newsletters, and spoke at in-person user group meetings. "There was no Internet back then, so you had to meet with somebody or talk to them to get solutions," Foster said.
A user community that grew up before the Internet has stronger links to innovation and assistance than groups that grew in the 1990s (Windows) and later (Linux), member for member. I like to think that every member of your Community carries several times more power and prowess than those from younger communities. As we've grown older things have changed a lot for the prospect of independent software and service providers. Yes, HP cleared out of your market. Its departure is even making companies like Cessna revisit how long they'll use the 3000 hardware no longer built by Hewlett-Packard. (There's a virtualization opportunity to replace HP's gear in the Stromasys product.) But HP's exit has also opened up the field for those Innovators and Improvers. Just look at how the world's change reveals itself in Eugene's survey of manager purchasing habits. One retired relic of that market: The Single-Vendor Shop.
Many HP customers have an almost blind loyalty to HP. In my years as an independent vendor, too often have I heard "sorry, we don't buy third-party products." This attitude, although sometimes justified by the desire to have a more easily supportable system, is usually quite incorrect because it deprives the user of the many advantages that can be derived from independent vendor products. However, condemning it won't make it go away, and every third-party vendor must live with the fact that a substantial part of the HP3000 market is forever barred from him.
Forever turned out to last less than 30 years. The change in the third-party vendor picture, whether selling software or services, has delivered a brighter opportunity for anyone who wanted to buy from more than HP. If an application enables your company to "change at the speed of thought," then the exit of the system vendor won't inhibit the useful lifespan of that application. Now there's only two parties in this ecosystem -- you, and anyone who can enhance and support your speed of thought. The third parties have become primary players with HP's exit. Since they created their places with innovation and improvement, I prefer to to call them independents -- or indie vendors, to borrow a term from the movies. The studio system isn't turning out as many great releases 30 years later, in either cinema or computing.
February 07, 2012
Managers report on mobile access to 3000s
Put a problem or a possibility in front of HP 3000 veterans and they will share what they know about solutions, usually on the 3000 newsgroup and mailing list. As we first noted last week, the problem of connecting the iPad or iPhone to a 3000 -- or the possibility of enabling this most mobile of clients -- sparked some tests and suggestions from your community.
"I've had a couple of requests from sales people wanting to log on to the HP 3000 to do lookups," said Randy Stanfield of Unisource. It's a company using the HP 3000 in support of its business selling printing materials such as papers, facility supplies and equipment, and packaging materials and equipment.
Telnet, as we noted yesterday, is the state of the art for apps to communicate with the 3000. A telnet client will most probably not know anything about HP escape sequences, so the app access will be nothing more than character-mode.
Consultant and security expert Art Bahrs reports he's found a couple of telnet emulators, and wondered if WRQ might have one that runs on iOS. Alas no, and WRQ became a part of Attachmate years ago. Its Reflection line still offers NS/VT and telnet links to 3000s. Attachmate has no iOS apps, a fact that's easy to confirm because the Apple App Store is the only source of apps that don't need a jailbroken phone or pad. Jailbreaking adds power and options to these devices, but deploying jailbroken iPads to a sales force is a strategy that can change a career.
Then Bahrs checked back in to report on zaTelnet v 3.3, from zaTelnet. Bahrs and other 3000 vets are running tests to see if an iOS device can manage a 3000, access that's a few steps short of user-grade interfaces to 3000 applications.Bahrs said that he was able to test the free version of the telnet iPad app zaTelnet. Many apps are free in this category, with a more fully-featured complement for a few dollars more. "It definitely would work for a quick and dirty trouble shooting session, or to check on a job, or support a user with an abortio or such," he said.
Security is another testing point. ZaTelnet is a SSH2 client for iPads, iPhones and iPod Touches. It emulates terminal VT100 and partially xterm -- enough for console programs. ZaTelnet supports SSH2 authorization by plain password, interactive password and private key file.
"SSH is an option," Bahrs said of the secure shell handler on ZaTelnet, "and it did work successfully both ways. There are times that good old fashioned telnet really does come in handy when doing testing, so I test both."
Mocha Telnet, which we mentioned yesterday, gets the job done for 3000 management. "It works perfectly on my iPhone," said support provider Gilles Schipper, "even the Lite (free) version. I can even run HP Glance on it. While it doesn't look too pretty, one can decipher the output. I set the termtype to "hp," rather than the default "vt220."
January 23, 2012
Quality, emulator futures slowing migrations
Some of the migration tool and service suppliers are expecting migrations from the 3000 to slow to a trickle this year. Alan Yeo of ScreenJet told us last week that the chance of extending the life of 3000 applications, by using the Stromasys HPA/3000 software, is going to put on the brakes for the sites that didn't have a clear future strategy for their 3000 servers.
Even without the possibility of replacing Series 900 hardware with the PC hardware plus software that starts at $15,000, most of the 3000 programs in production are not broken. They continue to do the job they were built for, although they could work faster, or connect better to new peripherals.
3000 managers wonder about these things. "Am I the only one out here?" they ask, in public forums like the 3000-L mailing list and newsgroup. The answer is no, you're not the only one out there. In fact, the populace of 3000 customers is surprising, both in its numbers and the work these systems do. The brand-new chairman of the Connect user group's board has managed a 3000 shop for many years. Steve Davidek is on the record as a fan of the 3000, even while his shop at the City of Sparks has migrated numerous 3000 apps.
"The City of Sparks, Nevada will be running an HP 3000 and BiTech Payroll at least through 2011," he said on the newsgroup. "Maybe longer, as the process to convert to our new system was hampered by the amount of budget we are allowed. Then again, why the rush? It is still the best out there.""We have three-year migration plan," said another manager on the 3000 mailing list. "And I doubt my last three HP 3000 shops have a plan in place yet."
Brett Forsyth, a reseller who has more than 20 years experience in the 3000 market, did a survey of his client list and found almost 200 HP 3000s still active within the last year or so.
My number totaled 197 known active. These are HP 3000s in a range running from Micro XEs to N-Class 750 4-way systems — and those are just the ones that I know of personally. Keep in mind that some of these clients have multiple units in multiple locations.
This is the beast that just won't die, in spite of Carly [Fiorina] and all the other MBAs who thought they knew better.
The largest best-known installation might be at Navitaire, the airline billing company. Mark Ranft last reported that the enterprise which was once known as Open Skies runs more than a dozen of the largest HP 3000s that Hewlett-Packard ever sold.
We have 21 HP 3000's. Eighteen of them are the largest, fully-loaded N4000-4-750 systems you can get. We have migrations to Windows in various stages, but there is also a very real need for legacy data access after the migration. The alternative is to migrate all the data and all the archival history, and that can be costly.
These are the sites that Stromasys will hope to attract with an emulator, pushing the horsepower of every emulated system to a minimum of an A-Class, with a top end even higher than those 750s. For those companies and organizations constrained by budgets, the goal of maintaining "the best out there" may dictate a lull in the march to migration. There's nothing wrong with a lull. It gives the companies who help migrate data, or servers, a chance to polish their products and collect a few more reference success stories. We're glad to spread both kinds of stories -- emulation as well as migration -- so get in touch with us as you or your clients make news.
January 12, 2012
Oracle serves Nordstrom's ex-3000 users
Ecometry e-commerce software once landed the HP 3000 some of its best-known customers. Store chains seen in shopping malls were also among 3000 owner ranks. One of the best-regarded retailers has replaced Ecometry when the company shut down its HP 3000. Both catalog and online sales were changed at Nordstrom. The beneficiaries of the multi-step move at Nordstrom were IBM and Oracle, but HP didn't get shut out completely.
Bob Thompson of Nordstrom's Sales Systems group said that the company's HP 3000 was used only for Ecometry processing. Triggered by HP's pullout of the marketplace, the retailer started to re-evaluate its e-commerce software along with the platform. Ecometry lost out to Sterling Commerce, a software provider which has become part of IBM. The software is listed under IBM's Selling Solutions.
Thompson said the company converted all of its Ecometry data to Oracle. The Sterling solution is running on Java, Oracle and Linux, but Oracle isn't a complete winner: There are a few Solaris boxes waiting to be replaced.
The migration started with a new COBOL program. Nordstrom wrote one to read its IMAGE/SQL data and convert it into XML. Then services were developed to use the Sterling APIs to add the data to Sterling's Oracle database. Nordstrom converted two years worth of data to import into the new software. Then even more Oracle embracing commenced. Nordstrom was not rewriting or doing a lift and shift migration. The strategy called for an application replacement and data migration.
"This part involved a direct conversion of all the Ecometry IMAGE datasets, for all time, to Oracle for historical reporting," Thompson said. "The initial COBOL conversion effort took close to a year," Thompson said, "plus another four months for the second part" to create the historical reporting facility. After leaving behind the HP 3000 and Ecometry and IMAGE, HP was left with a hardware assignment to fulfill at the company. But the Business Critical Systems, running Itanium, haven't been tapped from the HP product lineup."Originally, Solaris hardware replaced the HP 3000," Thompson said, "but now most of it is Red Hat Linux on HP hardware. Ecometry was all that was run on the HP 3000 here; warehousing has been done on the IBM AS/400, and other functions were done on IBM mainframes. But now all that's mostly Unix/Linux."
Choosing Oracle wasn't really the leading option in the decision process. Sterling's software uses the database. This is a typical way for Oracle to extend its reach into a migrating HP 3000 shop. Oracle is everywhere, but so is its reputation. Jeff Kell, working at the University of Tennessee at Chattanooga where HP 3000s are installed, says Oracle's got its baggage.
"Oracle is more of a third-party-requirement than an elective," he said. "If our applications ran on SQL Server --- for the ones that do, we typically take that route -- or MySQL, or anything else, we would avoid the pain."
But that kind of requirement brings in a database that can soon act as important as any operating system. Eloquence database creator Mike Marxmeier has said, "Being an Oracle user is really like being a user of an operating system, not a database."
Although Oracle now has a complete database-OS-hardware stack, just like the HP 3000 does, Kell said there's not much attraction for the Sun OS or hardware. "We have a similar aversion to Solaris for many of the same reasons as we now have an aversion to HP platforms," he said. "Java is unfortunately inevitable and something we do accept, even after the Sun/Oracle transition -- much as we still use HP's printers."