December 30, 2005
Where the future might lead, next year
This is the time of year when editors try to sum up the 12 months that have passed. In many cases, the publishers of IT news sources have little news to digest at the end of December. That certainly hasn't been the case for HP 3000 customers or this newsletter and blog. We got what appears, to some, to be a game-changing announcement last week: HP support for the 3000 now runs at least to the end of 2008.
However there's some debate over the details that might lead you to see a different future than "Wow, now we can use our 3000 longer than we thought!" (You can, by the way, provided that you can get along with basic, reactive support from HP).
Among those debated details are the ones omitted from HP's Dec. 20 report, according to OpenMPE chairman Birket Foster, who also happens to lead an HP e3000 Platinum Migration Partner (MB Foster). He shared his thoughts about source code licensing timelines — or the lack of them — in a message to us.
"The lead time and technical resources required to make the turnover of the code were not addressed in this series of [HP] memos," Foster pointed out. HP's only comment about source code turnover to a third party came during the OpenMPE meeting in August. Mike Paivinen said that HP believes turnover of the source — said to be in millions of lines long — to a third party will take less than a year.
It's important to note that HP's source code plans, as defined in its Dec. 20 communications, do not provide for all of the MPE/iX source to be licensed. From the FAQ:
In the future, when HP no longer provides MPE/iX support, HP intends to license major portions of MPE/iX source code to qualified providers for the purpose of helping them support their customers. Providers may be able to use MPE/iX source code to investigate solutions, develop work-arounds, and create binary patches to solve customer problems. Licensing terms will be determined at that time.
Foster noted "there are no thoughts or comments on what 'qualified provider' will mean." But we expect to know more next year about what kind of company HP wants to do business with in the 3000 support space. The vendor has said that third parties will play a role, at some point, in the delivery of HP 3000 support with the HP brand on it. (Go ahead, look it up in our 2004 report from the final HP World. Bob Floyd, VP of Americas Support delivery, made the comment. Floyd retired this fall, however, taking the Enhanced Early Retirement offer from HP. But we digress.)
Finally, Foster points out the source code release gives "no definitive timeframe, which makes it hard for anyone to build business plans around! But that may have been the purpose."
It's easy to see that the future of OpenMPE has to move toward becoming a user group in 2006; the project to create a virtual MPE/iX lab has been set off by two extra years. At least. We'll be here and in print to keep up with the details, debated or not.
We'll see you back here on Tuesday, Jan. 3, following the New Year's holiday. Our very best wishes for your 2006 — a year that doesn't mark the end of anything, as far as we can see.
December 29, 2005
Alternative boot up strategies: internal disk
HP says that customers are moving away from parallel SCSI-attached storage for their HP 3000s. At the recent HP Technology Forum, Jim Hawkins of HP's MPE/iX IO labs said FiberChannel Storage Area Networks and shared tape libraries are becoming popular. The HP 3000 supports SAN from the XP series of RAID devices to the VA7100 disk arrays.
But how much should you rely on a RAID or SAN device? Internal storage devices might be yesterday's tools, but the modest drive inside your HP 3000 can still be very useful, even if your company has invested in the FiberChannel storage solution of the VA7100.
Moving to the VA solution has great benefits. Last year we interviewed Donna Garverick about the use of the VA7100 array with the 3000s at Long's Drug. But booting directly from a VA array — well, you'd better have an N-Class server (native FiberChannel installed) or a very expensive HP A5814A-003 Fiber/SCSI router (if you can find one) for the 900 Series servers.
The Crossroads SA-40 Fiber/SCSI switch will link a VA array to 9xx 3000s. It just won't let you boot your MPE/iX system from any of its drives. Crossroads isn't interested in adding this functionality to its less costly array connectivity solution. Craig Lalley of EchoTech recommends the affordable Mod 20 arrays for boot capability
So if you're considering a move up to the VA arrays for your 9xx HP 3000 — or even to the XP line of HP arrays — your internal drives can remain as important as ever. Enough to even duplicate them, according to Garverick.
Garverick noted that a second bootable disc inside your 3000 can take some forethought, but it's essential in the event of an LDEV 1 failure. James Killam of HP reported to the 3000 mailing list, "Keeping a bootable image of MPE on one of the internal drives... saved me once at 2 AM when we lost total connectivity to the XP array the system was attached to and we had some serious troubleshooting to do."
I do have one internal disc and it is expressly for memory dumps. A second internal disc with a bootable image would be wonderful insurance. It would take some fore-thought to be able to manage it all... but there’s no reason why you can’t have two bootable discs. I’ll point out the obvious: if LDEV 1 is internal, and you have a multi-disc system volume, and the remaining system storage is on a disc array — uh, what’s the point? If LDEV 1 fails... you’re toast!
She gave a report on Long's VA7100 maintaining 3000 uptime as promised. "We actually had a drive fail on my VA array. The array worked perfectly and switched over to the spare without a blink. Given that I’ve got two systems sharing this array, I am more than pleased with how well it worked."
Drive failures are among the most likely of hardware problems you will encounter using an HP 3000. Considering how inexpensive disk devices have become, a second internal drive in a system can make a big difference in recovery time.
December 28, 2005
The size of your ecosystem
With HP's news of a support extension, we can't help but recall the vendor's advice about your ecosystem, circa 2001. Back then, marketing manager Christine Martino was certain that the 3000 ecosystem wouldn't be a safe place to habitate in say, five or six years. Which would put HP's range of safety at late 2007, based on its estimates from four years ago.
As we all revise our calendars for an ecosystem with "at least" another full year of safety, we ponder the question of the size of your ecosystem. HP insists that the majority of customers are at least planning a migration, if not already underway. How many are still using their 3000s is a favorite question to kick around.
The last analyst census of your community was in 2003 by IDC. Two years ago, that company estimated between 12,000 and 16,000 HP 3000s were still running. Do whatever math you like; it still looks reasonable to us to say that "at least" 8,000 HP 3000s are running today, worldwide. IDC estimated that the installed base dropped from 23,000 systems in 2002.
VP Jean Bozman of IDC said in an interview after HP's news that the support extension might prompt a new IDG estimate of your installed base. "We're now going to feel like we should come up with a new estimate," she said, adding that she'll report the results to the NewsWire. Bozman added she was surprised to learn that HP will deploy the resources to extend your ecosystem an additional two years.
"I didn't know [the customers] had more time to have this kind of service and support with some level of commitment of revenue from HP," Bozman said. "It's a good sign that HP isn't going to walk away from these guys."
December 27, 2005
Holiday week chills response to HP extension
Since the news surfaced just three days before the holiday weekend about HP's extension of 3000 support, users, partners and customers have not sounded off about the change in their calendars and plans. But some have sent us short reports we can share in this quiet week for technology pros.
Remember, this week was supposed to be the last late-December you could enjoy as a 3000 owner if your company relied on HP's support. Now that will be December of 2007. Herman Schweiker said the HP extension wouldn't have any impact on Air Frame Manufacturing and Supply. "We've already moved on to other support providers" describes the situation at the overseas technology exporter.
Jeff Kell, the curator of the essential 3000-L mailing list as well as the OpenMPE list server, as well as system manager of the University of Tennessee at Chattanooga's HP 3000s, said the university has already moved its third-party apps off its 3000s, but the custom-written applications remain in place. Kell said he's relieved for the extension, but not pleased HP will hold up its MPE/iX source code license another three years.
"I'm very relieved that support (SW and HW) will be available, even though we will likely transition to third party anyway," Kell reported. "It is a little reassuring that 'legal' hardware replacements and software fixes will be available. We aren't planning any new development and could live with the current software situation without enhancements. There is the matter of MPE keeping pace with new peripherals, but they seem to have coverage for the immediate future."
"I'm very annoyed that HP is stalling third party licensing of software/OS code. The two-year delay put the nail in the open source coffin; the OS will be incredibly stale by the time any outside access materializes."
About the applications future at the University, Kell said, "We are left with our in-house legacy student information system, some pieces of which are still circa 1977, when development began. We don't have the staff or time for a significant redeployment or rewrite, and the software life cycle of the system is on it's last breath anyway. Replacement will be a "forklift." Choice of a replacement has been stalled indefinately pending "system-wide" studies of various alternatives in an effort to make a "system-wide" migration (each campus currently has their own application systems). Our choice will likely be mandated regardless of our wishes."
HP's announcement won't have an impact at Dynacraft, a maker of heavy-duty truck parts. A third party services the company's 3000. "Ironically, our system replacement project got funding less then 24 hours after HP’s announcement," said Craig J. Hoppler in the IT group. "Dynacraft is already receiving HP 3000 support from another vendor. The company is moving to an ERP package from SAP. We are a division of a larger corporation."
December 23, 2005
Listen to that holiday gift that gives for years to come
HP dropped off its holiday gift to 3000 owners this week, a topic we comment upon in about six minutes of our weekly podcast (6 MB MP3 file). Extending vendor support beyond 2006, to “at least 2008”, HP seemed like it had to admit that migrations were taking longer than anyone expected. Especially back at the end of 2001, when the vendor cut its 3000 business off with a five-year farewell. But the gift came wrapped in the colorful paper of transition success, somehow.
Have a listen to our holiday show, and have a safe and merry weekend. We'll see you back here for more 3000 news and views on Tuesday, Dec. 27.
Update: Those NDAs are still in place
HP's e3000 Business Manager Dave Wilde contacted the NewsWire to give us an update about the state of the non-disclosure agreements that have covered receiving 3000 support post-2006. On Dec. 20 we were told those NDAs' terms and limits would be waived, but Wilde tracked down the latest NDA advisory for these companies.
Wilde said about the announcement to extend basic-level support to at least 2008:
As a clarification of an earlier statement made by HP to the NewsWire, HP indicates that this announcement does not invalidate any NDA covering previous HP support discussions with customers. HP also indicates that it will work directly with customers who had previously approached HP, to reconcile the new basic reactive support offering with any prior discussions of support needs.
We'll have a podcast with our holiday analysis of the announcement posted up later on today. Our regular blog entries, with more details on the extension of HP's 3000 business, will resume on Tuesday, December 27.
December 22, 2005
The unique value from HP's extension: labs
HP surprised some of its customers by announcing on Dec. 20 that its 3000 support business will end two years later than 2006. Some debate has emerged about what this means to the community, to OpenMPE, and to the vendor itself. OpenMPE had a conference call today to discuss its fate in the wake of the HP decision to withhold source code for an extra two years, at least. If the advocacy group is to survive, it might need to be as a user group, rather than an organization running a lab.
The OpenMPE directors will decide what's best for the organization's future, but now the 3000 community's future includes something very unique: an extra two years of work on MPE/iX by its creators. Bug fixes, probably, make up the biggest list of 2007-08 projects. But now the labs in Cupertino and in Bangalore, India have more time to get those many patches out of beta test and into general release. HP reiterated in its letter to customers that the vendor wishes more of you would test those enhancements — the ones which fewer than 300 people voted on in 2003 and 2004 System Improvement Ballots.
The labs have other work to finish, or even succeed with for the first time. The IMAGE/SQL LargeFile database mess is still in repair. HP had a patch for these datasets to replace jumbos, datasets that have caused corruption so routinely most partners advise customers don't use them. HP moved out the repair patch into test with a third party vendor this year, but that patch didn't make the grade.
We've been told by database experts that LargeFile support is so complex it can have plenty of negative performance impacts if it's not done with the most elegant of engineering. Elegant engineering is especially hard when the test subjects must be massive datasets, which take days to test with all but the hottest of HP 3000 hardware. Remember, these are servers that are hampered by slowdown code in all but the biggest models of N-Class devices.
These are the kinds of problems the vendor's labs are best equipped to solve, and now there's more time on the clock to do so. What HP's 3000 lab will do for you during these two years remains to be seen.
What the vendor wants us to remember is that support from HP is not a long-term solution. The market understands that. Otherwise, people wouldn't be moving to third parties so often for support. Although this announcement was about HP's support, it's not the call-and-fix kind of repairs that the customers need from HP. Third parties have handled that kind of work in able fashion for many years, at a serious discount.
No, what the 3000 community needs is the effort of HP's engineers, the factory-level experts, to resolve things like the LargeFile mess. Since HP won't be dishing out its MPE source code for at least three more years, no third party is going to be able to amass factory-level experience until sometime in 2009. Unless HP's 3000 engineers can be hired away from the vendor, HP has cornered the market on patches.
We might think of that as an extension of what the customers are comfortable paying for. After all, patches are free to everyone. It's good that HP will be running a support effort for an extra 24 months to keep writing them. The tough nut to crack: getting customers to test them. Maybe with the extra time on the clock, some sites will be willing to take the risk and test patches. This announcement was about good will, too. One customer called it "a wonderful Christmas present." While many partners don't see it that way, it's easy to understand how two extra years will look like HP is listening, and responding, to customer timelines.
December 21, 2005
Something new: more HP lifespan
It took less than six hours for the 3000 community to digest yesterday's announcement of extra 3000 support — then a reader wanted to dismiss the announcements as "Nothing to see here. Move along."
We want to disagree. I had the chance to ask 40 minutes of questions just a few minutes ahead of HP's release of its announcement. I spoke with Jim Murphy from HP Services — ultimately the arbiter of how long HP will remain in the support business — and e3000 Business Manager Dave Wilde. Unlike what that 3000 customer had to say, there is something to see here. The question for any 3000 shop is, “Will I have any use for what I see?”
HP will remain in the support business in 2007 and 2008, but it will be “basic reactive” support, unless you need mission-critical enterprise level support. Basic reactive gets you HP’s repairs, but nothing proactive. And the vendor’s “6 hours from call to completion” guarantee isn’t part of the basic reactive service, according to Murphy.
HP's FAQ file on the extension of its 3000 business says, "Based on local capabilities, HP will offer at least hardware maintenance, software update services, and business hours software technical services on selected components."
As for those limitations and exclusions, those will apply more to the mission-critical level of support. If you’re way out in the boonies, or HP’s just not servicing much of anything in your geography, then mission-critical is either going to be A) More expensive than it’s been, or B) Just not available. Prices change after the end of next year, or the end of your service contract, for mission-critical.
HP claims it’s going to keep resources (read: MPE experts) in place to ensure a satisfactory level of support. Murphy even said HP is training its people (in, I presume, MPE/iX and the HP 3000).
For most customers, asking HP to extend their support contracts is all it’s going to take to get the vendor’s support for two extra years. No proof of migration plan, no promise to buy HP systems to replace your 3000. If you want them to support 6.0, forget it. We didn’t hear if they planned to drop 6.5 anytime soon. Probably not; still ample contracts out there, and 6.5 is pretty stable. There’s still no plan to let 9x7s boot up on 7.0, no plan to drop the slowdown code that cuts all A-Class and most N-Class server performance to a fraction of HP-UX counterpart systems.
But this delays HP’s exit from the 3000 market by two years.
Five years was not enough to let HP leave this market without leaving revenues on the table. This move continues a profitable revenue stream for HP — at the same time that it gives customers more time to either set up an alternative support provider, or, as HP insists is the best course, get off the server. HP was so adamant about this being the take-away message that it started off its Media Fact sheet with two paragraphs of how the majority of its customers are in some stage of migration, and that the best plan is get off the 3000. The fact sheet noted this before revealing the news, pretty unusual structure for a communique to the media.
HP also said that those customers who signed an NDA promising not to reveal their support extension now have the NDA's provisions waived, since HP made the blanket extended support offer. So the wraps are off to talk about their support extension experience.
Whether basic reactive support is enough to keep a business relying on an HP 3000 is up to the customer. HP says it found a better stream of parts available for the 2007-8 period than it first expected. And it still considers third parties to be a potential part of its own service supply chain for the HP 3000. For the moment, however, the HP support you get is going to come from an HP employee or contractor.
Third party support actually now takes a step up in a comparison with the new ‘07-’08 levels of service. Most companies offering support won’t charge as much as HP to do mission-critical support. What they will have, in their toolbelts, is the HP support backup that was going to disappear in just 12 months.
As for the third-party MPE licensing offer, it’s real, but it’s hard to say when it will be extended, or to who. Or what will be in the license. HP's said, "HP intends to license major portions [italics ours] of MPE/iX source code to qualified providers for the purpose of helping them support their customers." Right now HP doesn’t have to open up the source code to anybody until December, 2008, when the vendor is currently scheduled to end all its HP 3000 support. It could be later than that, according to HP. They say they keep listening to what customers want to keep buying (if you overlook the fact that the customers wanted to keep buying 3000s in 2001 -- just not enough customers to keep HP interested in building them.)
Oh yes, definitely something to see here.
As for the relative silence from the customer community, this might be the result of making an announcement three days before the Christmas holiday weekend. Much of the world is already making plans or departing for R&R. As for the business planning of the 3000 sites’ budgets, well, 2006 is already spoken for. All this does is change the options for 2007.
It’s too bad this decision didn’t come at a time when more people were listening and allocating budgets. But HP did more to announce this than its last two updates to OpenMPE requests. In those instances, responses came in the form of postings to this mailing list and the OpenMPE list. This time out there was PR support and an outreach to business analysts and the mainstream IT press. You’d think the vendor had something to sell here, like goodwill in a holiday season, or another couple of years of support.
December 20, 2005
HP extends its 3000 support date to 2008
HP released news today that changes its timeline for ending its HP 3000 business. Now, 2006 means at least 2008, and perhaps beyond.
In a surprise move, Hewlett-Packard has extended its support for the HP 3000 beyond the end of 2006, offering customers what it calls "basic reactive support services" for MPE/iX through "at least" the end of 2008.
HP issued a media fact sheet and sent a customer letter to its installed base to announce what amounts to an extension of the vendor's 3000 business for an additional two years. The announcement, released on Dec. 20, also included a statement that HP intends to offer a license of MPE's source code "to one or more third parties." HP will offer that license only once the vendor stops offering "services that address the basic support needs of remaining e3000 customers."
But the adjustment of the 3000 business roadmap emerged as the unexpected element in a long-awaited announcement. HP began its announcement with a statement that a majority of customers and partners in the 3000 community have either completed migrations, or are actively working on those projects. HP's extension of its 3000 business came with multiple caveats, a warning that support prices from HP for the 3000 would rise — and another chorus of "transition to a new HP platform as soon as possible."
The realm of possibility for its 3000 customers may have prompted the HP extension of its MPE business. HP wants to give the customers making a transition to newer HP solutions "a little extra time and safety margin." Many customers have found it impossible to meet the previous December, 2006 migration deadline.
"HP recognizes that some companies may need to run their e3000s after HP’s current end-of-support date due to business constraints, transition timelines, or to retain access to data for archive or regulatory reasons," the fact sheet stated.
HP Platinum migration partner Speedware reported that the support extension is a deal that's been offered to HP 3000 sites "for a little while" — provided that the customers sign a non-disclosure agreement about the extension.
Speedware's marketing director Chris Koppe said the HP support extension doesn't change much for his company's migration services and tools business.
The later deadline has been in place for any customer who asked for it in support of a migration project. Koppe described the offer as one that "has been made available on an exception basis to specific accounts where their migration would take them more time than was available.”
"We were telling customers who were deep in the sales cycle and willing to sign NDAs for quite some time now," Koppe said. "I'm not sure of the number of customers that HP already has signed up, but it's a lot more than I thought."
MB Foster, another North American Platinum migration partner, said the offer of extra support was "one of the worst-kept secrets in the marketplace," according to founder Birket Foster. The extension of HP support doesn't change the business model at Speedware, or MB Foster, according to their officials. But offering basic level reactive support won't meet some customers' needs, Foster added.
While some customers will welcome the potential for more time to migrate, Foster said the HP announcement is introducing some confusion among others. "We had a customer who looked at this and said it would not be enough to make them supportable — but their senior management felt they could take the extra time," Foster said.
December 19, 2005
HP peers back into its past
Hewlett-Packard, one of the biggest technology companies on earth, began its life inside a garage in a quiet professor-laden neighborhood of Palo Alto, California. In 1987, the company's Ground Zero gained historical landmark status with the State of California as "The Birthplace of Silicon Valley." This year HP is taking the time and spending the money to refurbish and renovate The Garage and the house in front of it at 367 Addison Avenue back to its classic 1938 look. HP now owns the property, a significant part of its history, once again.
This week, while we wait for HP to make its promised "second-half of 2005" announcement about the post-2006 era of the HP 3000, we suggest you take a look at the new video HP's put together about the company's roots and that garage. HP has been through a lot of changes since 1939, when Bill Hewlett and Dave Packard sold their first oscillator for $847.50. But no matter what is announced this week before HP goes on its year-end holiday sabbatical, looking back at the company's heritage makes it easier to believe the HP Way that spawned MPE still lives in the intentions of some of the HP employees who still serve the 3000 community.
Have a look at the film at the company's Garage renovation Web page. (The link is on the right-hand sidebar of the HP page). Stay tuned to this space for an update — whenever HP makes it — on the source code decision and whatever else the vendor has to say about helping its customers preserve a heritage. History is something this large corporation obviously still considers important.
December 17, 2005
Are there support changes coming?
HP 3000 customers still want updates about HP's support of MPE and the HP 3000. We get messages regularly from people like Alex Purves, the Six Sigma Expert at Raytheon's UK operations, who asked last week: "Can you please provide me with the latest position of HP support of MPE/iX?"
This question may have a different answer very soon. We don't know what HP's end-of-2005 announcement will cover, but a pretty safe bet is an update on the MPE source code licensing issues. HP has not promised any kind of third-party license for MPE post 2006 — yet. But time draws near for the decision-mulling process to end; HP promised news by the end of this month.
To answer Alex's question, at the moment, HP is committed to support MPE/iX with HP engineers until Dec. 31, 2006. All year long HP has said, "2006 means 2006." We'll see if it that's what it means soon enough. If HP means to shift that deadline, the sooner the vendor announces that good news, the more it will seem like a goodwill gift to a customer base that has been struggling with migration deadlines by the end of next year. We can't say if any change in that deadline is even a remote possibility. We can only say that it's not impossible — because it's not the end of 2006, yet. And HP has been listening to customers who are migrating. The vendor knows there's customer anxiety about meeting that 2006 deadline.
In the meantime, there's other questions to consider, like, "What is HP's support level, anyway?" From reactive to proactive to mission-critical, there's lots of levels to spend support dollars at HP. Reactive works for most people: Something breaks, and HP helps you fix it. Other levels of support don't get nearly as much pickup by the 3000 community, a very DIY bunch that's been cautious with their spending.
Across the board, you can spend less and get better support for most enterprises with third parties. The level of HP support to help out on the most complex problems is about the only place where HP might surpass a third party — if HP still has the bandwidth to write patches to fix problems like the LargeFile mess in IMAGE. (We haven't heard anything about that fix going into beta-test, five months after it crept into alpha testing.)
The customers' spending caution is the reason those migrations have moved so slowly so far. Slower than HP expected. Slower than third-party partners expected — both those selling migration services and tools, as well as those trying to get alternative support businesses ramped up. Companies haven't budgeted much for their 3000 enterprise, compared to Windows or Unix installations. Migration is big and can be costly, too, depending on which strategy you choose for your transition. People have expected 2006 to be a watershed year for migration — at least HP was pointing to that kind of future just two months ago. That's when HP's Alvina Nishimoto, manager of the HP migration center, told HP Technlology Forum attendees that HP expects something of a panic in 2006. Back in October we reported:
According to HP’s Migration Center manager Alvina Nishimoto, “The majority have some sort of plan by now,” she said. “It’s going to be a little bit of a panic now, as people start to wake up to the fact. The ones that are waiting the longest are the packaged app folks, because they do think it’s going to be faster [to move to a packaged app]. It’s faster, but they still have all their surround code, and they don’t necessarily think of the implications of the surround code."
If the vendor can avert that panic, somehow, thousands of customers might feel some relief. HP has been saying for quite some time now it will extend some support services beyond 2006 — where it can, as resources allow — if you have a clear migration plan for them to review. Just ask customers like Hertz what their migration deadline is. If all that doesn't exactly sound like "2006 means 2006," well, HP can reply that migration is a complex matter. It must be, when we see how long it has taken for a decision about source code licensing to surface. If it surfaces soon. Stay tuned here — we'll report what HP tells the world.
December 16, 2005
Wait for HP's answer, watch that clock
In this week's podcast (6MB MP3 file), we talk and listen for about seven minutes about changes to the HP 3000's future. Change happens, so be ready for it. That’s the mantra HP has repeated for years, especially to its enterprise customers. Slick TV commercials showed a business morphing. Those computer graphics are not the only thing that’s likely to change from HP. The answers to the questions that have been asked by OpenMPE might get unexpected answers. We don’t know yet. But let’s entertain some possibilities while we wait this week.
December 14, 2005
Consider data synchronization during transitions
By Nicholas Fortin
For readers who own an HP e3000 server and are interested in the subject of transitions, data synchronization is not talked about very much. One of the most critical topics to focus on during migration planning is the deployment strategy, especially its impact and relation to data migration and data cutover for moving to production. Sometimes you will need to use data synchronization technology to support your desired deployment strategy.
A database synchronization technology facilitates the synchronization of data from one database type to another, usually located on different platforms (making sure updates performed on one database are also performed on another database, as well as keeping the two synchronized). To use a basic example of how this would apply for IMAGE to Eloquence, you migrate an IMAGE database to Eloquence, and then keep the Eloquence database synchronized (updated) with the transactions performed on the HP e3000. Why would you want to do this? Well, I have some examples below.
First, let me go over some migration challenges or caveats related to this topic. You may need a data synchronization technology if one of the following scenarios apply to you:
1. Your databases are large and your available downtime small enough to make it impossible to migrate all your databases in the final cutover period (this is sometimes referred to as the big-bang approach). To explain, first let’s make some assumptions. We’ll assume that the databases are large enough to require a significant amount of time to migrate from one type to another (we also underline — although it’s probably implied — that when the databases are migrated, they need to be put offline so no transactions are performed in order to migrate a specific snapshot of the data).
In addition, we’ll assume that a business has a small user downtime period allowed for switching the users to the new system. These two traits combine to pose a data migration challenge because at the moment of final data cutover, before moving all users to production on the target system, the time needed to migrate the IMAGE databases surpasses the time limitations for the HP e3000 system databases to be un-accessed (i.e., database downtime). These two criteria show potential synchronization need.
In this scenario, once a snapshot of your databases is taken (using a backup, copy or an export of your databases) during the transition project, you could use a synchronization tool to keep your target database synchronized with updates continued to be performed on the HP e3000. This scenario represents unidirectional synchronization from source to target.
2. Your testing strategy requires platform parallel testing of screens and reports (and sometimes transactional programs) — testing whose results depend on the exact same data values from both source and target databases at any given time. This scenario represents unidirectional synchronization from source to target.
3. You envision a contingency backup plan to revert back to the HP e3000 after having moved into production on your target server (for a set period of time like a week or two), in the event the migrated application environment has some serious problems. You would then need a synchronization technology to keep the IMAGE databases synchronized with the target transactions. This scenario represents unidirectional synchronization from target to source.
4. You envision a deployment strategy which requires a phased deployment of portions of your applications into separate bundles migrated, tested and moved into production. In this scenario, you have the same databases (usually some of your databases) on your source and target system accessed/updated simultaneously live in production requiring that databases on both platforms be kept synchronized with minimal discrepancy delay.
Due to the importance of having data integrity between databases on different platforms, this scenario also usually implies using a two-phase-commit mechanism to ensure data is posted correctly before the application process can move on. This scenario represents bidirectional synchronization from source to target and target to source simultaneously.
In fact, for two of our currently on-going large migration projects, this topic carried so much importance that it was included in a strategies/technologies validation milestone phase occurring as the first part of the project. (This phase has the objective to make the customer feel confident in some key technologies which may not be as proven as others, or which may carry unanswered questions needed for the transition project).
There are a few synchronization technologies available which support a number of database types including TurboIMAGE, each of them carrying their own strengths and “particularities.”
Another branch of this topic is data synchronization for flat or KSAM files. Although flat files are usually small in nature, a company that has hundreds or thousands of flat files to migrate can be faced with the same challenges noted above for databases. Becasue the migration process challenges and solutions for files are different than for databases, file synchronization is out of the scope of this article. However, Speedware possesses a unique and innovative tool that can migrate and synchronize thousands of flat files efficiently, quickly and with integrity.
In summary, synchronization can:
• Help with a phased deployment for the data cut-over
• Help with 24x7 or extreme uptime server requirements
• Help to run HP e3000 and Unix systems in parallel in production
• Help for a contingency plan in case the deployment is problematic
This information just scratches the surface of this topic. If you want to find out more about data synchronization and how it would relate to your environment and particular needs, you should talk to an expert who has this knowledge.
Speedware is an HP e3000 platinum transition partner which possesses a unique expertise in this domain based on real experience. Speedware has several data synchronization software and tools available.
December 13, 2005
Choosing the faster platform
HP reports that choosing a database is one of the most significant decisions a migrating site can make. In the same presentation HP's Kevin Cooper mentioned that Marxmeier Software's Eloquence has racked up very good performance numbers, according to reports HP's received from its migrating customers.
But which platform you deploy Eloquence upon can make a difference. In general, the customer base has praised HP-UX as a faster Eloquence platform than Windows, although that kind of generality makes technical experts cringe. Still, there's this report from a user on the Eloquence mailing list:
Eloquence creator Michael Marxmeier addressed these questions:
In testing our software on both Windows and HP-UX we have found that programs that access an Eloquence database take quite a bit longer on the Windows server. The Windows server we are using has two 2GHz processors and 2GB of memory; our HP-UX server has two 750 MHz processors and 2GB of memory. For example, when running a C program that does a serial read through the database it takes 26 seconds on Windows and 10 seconds on HP-UX. On the HP-UX server we have Enable IPC set to 2, when it is turned off the program takes 22 seconds.
My questions are: 1) The enable IPC option does not seem to be valid on Windows (option not included in config file included w/Windows version; when added to the config file it does not make any difference in processing speed). Is there a way to configure a database server on Windows to get similar benefits that the enable IPC option provides on HP-UX?
2) Is there any info available that shows what specs are needed for a Windows server (CPU, Memory) so that it gets similar performance to a certain HP-UX server when accessing an Eloquence database?
The IPC transport is currently not available on Windows. When we tested this initially, it was not critical for the Windows performance (but made a difference on HP-UX). A quick/simple test with the current version seems to support this.
I’m afraid this [Windows performance specs question] is not easily possible, as both the hardware architecture as well as the OS are rather different. For example, if the CPU architectures are different, the CPU frequency has little significance when comparing CPU performance. For CPU specific performance I would assume the SPEC benchmarks (www.spec.org) might be able to provide a rough overview. However, please keep in mind that IO bandwidth and IO performance might also be rather different.
I did some quick/simple/unscientific test comparing a recent HP-UX box against a Proliant DL360 (dual CPU, SCSI) running Windows. The results seem to indicate that for this test the performance should be roughly equivalent (for a small number of users).
John Pittman of Ryco in Australia chipped in his humble opinion about expecting HP-UX performance from Windows systems:
Sorry, but why would you ever expect a Windows system to be as fast as a UX box at anything? I don’t care how fast the Windows CPU is, or how much RAM it has, its just not going to get there. We run 200+ sessions and up to 40 background jobs on an MPE N4000, on a 1x 220Mhz CPU and 1GB of ram, with about 10 IAMGE databases open, some up to 4GB in size. Windows chassis just don’t have the IO bus speed to match a real computer.
December 12, 2005
Holiday contributions to MPE's future
OpenMPE continues to wait for HP to announce its decision about a 2007 and beyond license for the 3000's operating system. But the advocacy group staffed by nine volunteers has already announced a plan to create a lab, staffed with 30 experts working on contract, to handle anything that HP will be releasing. Setting up a lab to build MPE/iX with needed patches takes time — perhaps more than HP believes, to do it right.
That's why OpenMPE is hoping to start collecting funding for lab now, in advance of HP's decision. Earlier this year the group announced its pricing stategy for these lab services. While OpenMPE has contracted with an independent accountant to handle the funding finances, one of the group's volunteers is the primary point of contact if you're signing up. (For more details on that lab services plan we refer you to several NewsWire blog articles, as well as the OpenMPE PDF file of its slides from that August meeting.)
Sign up with treasurer Paul Edwards, who has the deposit slips for the OpenMPE account, with a check payable to OpenMPE, Inc.
Paul Edwards & Associates
1506 Estates Way
Carrollton TX 75006
Phone: (972) 242-6660
Donna Garverick, winner of this year HP e3000 Contributor Award from HP, sent a message to underline that the corporate funding of the plan shouldn't scare off individuals who want to help on MPE's future.
Now, before y’all faint -- this is a *proposal* for *corporate* membership -- something/someone with deeper pockets than you and me! For the rest of us... I’ll now direct you to our membership page: www.openmpe.org/membership.htm See? Zero cost! Free...free is good :-) (You are a member, right?)
However, free isn’t perfect. I think everyone has some idea of what it will take to run all those racks of equipment required just to build MPE. There are facility expenses, heating, cooling and so on. MPE isn’t a free ride today and it isn’t going to be tomorrow either. OpenMPE will take any amount of money you’re willing to send. $1.... $5.... $20... whatever. (By the way, OpenMPE is not a charitable organization...so the money you are sending won’t be a donation from a tax perspective).
The NewsWire's coverage on the OpenMPE plans to protect the future of the 3000 in 2007, a year when many sites won't be migrated yet:
Podcast, about eight minutes, on setting up the labs
Why OpenMPE matters
Why HP believes a year is enough time
How an emulator is essential to hobbyist licensing
What a new negotiator will bring to the efforts of OpenMPE
December 09, 2005
Preparing to move to 7.5? Check this list
HP 3000 customers have been advised by HP to consider the latest release of MPE/iX for their 3000s. This 7.5 version of the operating environment has been available for years; HP has already PowerPatched it - a process that consolidates patches into one superset - several times to add functionality and repair bugs.
But moving to 7.5 from 6.5 includes a few things to monitor. HP's 3000 peripherals expert Jim Hawkins recently advised a customer about volume management and IO changes. The counsel Hawkins shared looked like good advice for anyone making the change to 7.5, the last version of the MPE/iX which HP will ever build for the 3000:
The only “gotcha” with LDEV #1 suddenly having more space relates to the case where you have multiple volumes in your MPEXL_SYSTEM_VOLUME_SET. After the update LDEV #1 will look “empty” compared to other volumes and it will become the target of all disk allocations until its free space count is equal to that in the other volumes. (From 0-90% full across the set; after we reach 90% on all volumes the disk with the largest number of free sectors is the target). This applies for any MPE Volume set where you add a new disk (which is kind of what you’re doing).
In this case you might even see an effect on system performance in the short term — as XM and directory and VM and File system allocations will all be focused on LDEV #1. For this reason it is actually better to have LDEV #1 (or any master volume) be smaller than the other devices — it always gets the XM and directory traffic but, on average, will be less busy than the other disks for file system allocations so, on average, shouldn’t be a bottleneck. Your mileage may vary.
To prevent this kind of performance problem, after you’ve completed the 7.5 update activities and completed a full system back-up, you might consider "moving" a few (larger) accounts that reside on the MPEXL_SYSTEM_VOLUME_SET. One method would be to "STORE @.@.X,@.@.Y;;PURGE;DIRECTORY" and "RESTORE" from the same tape. When you RESTORE the files MPE/iX will spread them across all the volumes as evenly as possible.
Part of the Large disk enhancements, MPEMXT4, “smooths” the extent spreading algorithm a bit more to help prevent “hot” disks. Call HP for this beta-test patch.
There's been talk for a long time about a "parking release" of MPE/iX for the 3000 as HP wraps up its 3000 operations. 7.5 looks like the most likely candidate for that release: the HP labs have developed enhancement patches for that version first, then moved to the 7.0 and 6.5 releases.
December 08, 2005
Itanium: Not dead yet, but...
It's not exactly an obituary, but yesterday's story from CNET reporter Stephen Shankland sums up the 11 years of HP's Itanium project with Intel as "a cautionary tale." Yesterday we sat down with an old friend, Terry Floyd of 3000 ERP experts The Support Group, and we mentioned Itanium briefly. "Could it really be 11 years ago?" we both asked, noting that it seemed like a long time for a chip that hasn't set the world on fire.
As Shankland's story notes, HP and Intel were predicting a blazing future for their joint project way back then. First it was code-named Tahoe, then Merced. Finally the pair of technology giants got first silicon shipped, but the competition wasn't standing around waiting for the release. AMD, IBM and even Sun rolled out their 64-bit entries. Shankland notes that Digital's Alpha fell in the chip competition — but you might argue that the Alpha demise was more about Digital being acquired by Compaq, which was then acquired by the makers of Itanium.
Itanium has a future in the HP 3000 marketplace. Shankland's story reports that Gartner says four of every five Itanium servers shipped during the third quarter carried an HP nameplate. Itanium is the successor to PA-RISC — which is still outshipping Itanium, according to Gartner's figures. But back in the 1990s, CIOs and software companies were told to believe Intel would conquer all with the design that was born in HP's labs.
If there's a caution in HP's Itanium story, it might be this one: Shipping something late is never in line with dominating a market. Late chips don't run any faster when they arrive late. HP will prosper by its investment in Itanium, if you believe that PA-RISC is out of gas. Itanium is the only long-term platform for HP-UX, so that makes the chip's fortunes important to the 3000 community which is migrating to HP's Unix. Some said that HP did its deal with Intel 11 years ago to gain strength of numbers in massive market share. Itanium is unlikely to accomplish that goal, so HP is left with a secondary success: Getting to shut down its chip fabrication and design operations, so it can spend that R&D money someplace else.
December 07, 2005
Sizing for an A-Class migration
Migration planning includes trying to calculate how much Unix box you'll need to replace that HP 3000 system. While customers have given the newer generation of 3000s good marks compared to the 9x9 and 9x8 servers, one report we saw recently confirms what HP claims: Moving out of the hampered MPE/iX environment which HP slapped onto the A-Class servers will give you some performance room. It seems that those A-Class 3000 systems, just like nearly all of the N-Class servers, aren't running as fast as they possibly can.
The report was just the latest entry we've heard about how HP's held back the 3000 line compared to its Unix counterparts. HP has actually told 3000 customers this is a benefit — at least that's the way it's presented to a room of customers who are asking questions about migration. This time Duane Percox of QSS was answering the following question from a migrating user:
We currently have an HP 3000 A-500 140 2-way, running a client server application using an IMAGE database and ODBCLink as the middleware. Would a RP3440 2x1GHz PA8900 CPU with 4 GB memory, compare in speed and performance? We are going to convert the IMAGE database to Oracle. Just trying to get an idea.
The answer began with a warning: You might not have as much HP 3000 power to replace as you think you do.
Percox answered with a set of notes, based on his company's tests of COBOL applications on HP's Unix systems (details at http://qwebs.qss.com/cobdata.html) saying:
• Your A500-140 is not running at 140MHz as advertised by HP. Due to CPU throttling, that may or not be as good as could be. You are actually looking at closer to 72MHz for that A-Class
• Throttled A-Class boxes also exhibit interesting IO timing issues as demonstrated by some very astute folks who would know, given their intimate knowledge of everything MPE. Here again you might not have as big a box as you think you have.
• I haven’t managed a production Oracle system, but from what I have read you probably want more memory than 4GB for that system if it is going to be an app and database server. I would start at 8GB, and consider even more depending on the number of connects.
• Make sure you are getting a 2-way server. I would never recommend running a relational database server with less than a 2-way. And you might even need a 4-way depending on the number of connects.
• Make sure you connect that new server to your network with gigabit LAN.
• You didn’t mention disk subsystems, which have a big impact on database performance. And no mention of the number of database connects, which also has an impact.
• I find that moving from TurboIMAGE to relational is about a 10-12x CPU hit for the parts of the app that are managing the database. Since your app also spends time doing other things, you don’t necessarily have to have 10-12x the CPU, but it might be a reasonable starting point.
• You are wise to get a 3440 system so you can put two more CPUs into it if needed. Also, the more CPUs the more memory (a vicious cycle). I assume you are starting with 2x1GHz in that 3440. Once again I will point out: minimum config is a 2-way.
• The MPE TCP/IP stack is performance challenged, so you will see networking improvements when migrating to HP-UX.
• TurboIMAGE/ODBCLink isn’t a performance screamer, so you might be in for some pleasant surprises in the positive direction.
December 06, 2005
Use SED to automate text processing
Up on the 3000-L mailing list, a lively tutorial broke out yesterday on using SED, a stream editor built in the open source community. Since 2001 SED has worked on the HP 3000, thanks to Lars Appel, a former HP support engineer who ported Samba to the platform in the 1990s.
SED's main MPE page is on Appel's page at www.editcorp.com/Personal/Lars_Appel/sed/. (Editcorp is a company that consults in the HP 3000 community, among other places. It also works with relaying the 3000-L postings to the comp.sys.hp.mpe newsgroup.) It's an at your own risk download, but support is available through the 3000 community. Yesterday's 13-message volley proved that; the community heard from Appel on one of SED's blind spots, along with a workaround.
Dan Barnes, working on a problem he had to solve in his MM 3000 environment, asked:
The issue is incoming data from another platform that is being fed into MM 3000. This data occasionally has some unprintable characters, which of course wrecks havoc on the MM application when it is encountered. To address this, the user, using a cygwin (Unix-like) environment on their Windows PC, developed a SED script. When they test the script in the cgywin environment it works just fine. But when done on the target HP3000 (7.0 pp2) it gets an undesirable result.
Barnes added that "The user thought that because MPE/iX is Posix-compliant, that this should work." He explained his user created the expression
sed -e 's/[\x7F-\xFE]/*/g' < COMSHD > COMSHD1
But Appel noted that hex 7F thru hex FE portion of the expression isn't supported on the MPE/iX version of SED. It's a limitation of MPE/iX, but there's a workaround:
Not sure if the regular expression usage here matches Posix or GNU specs, but my guess is the "\xNN" format, that seems to indicate a char by hex code, doesn't work.
How about something like sed -e 's/[^ -~]/*/g' instead, i.e. map the chars outside the range space through tilde?
Appel then noted a nifty Web reference that documents the man pages (manual pages) for the 3000's Posix features:
Contained within this resource is the documentation on how the 3000 handles regular expressions — the rules, if you will, on how to form compliant SED expressions.
Oh, and the solution took only four hours to deliver to Barnes. Not bad for an "unsupported" part of the 3000 experience.
December 05, 2005
Keep an open mind: don't be scared about migration
In our podcast for this week, (7MB MP3 file) we talk for seven minutes about migration off an enterprise platform — a big enough job for nearly every 3000 shop. For some, the task is the biggest thing they’ll ever try to accomplish in computing. Even bigger than Y2K. Six years ago there was no bigger scare job than Y2K warnings. Those kind of stories get some shops onto their feet, sure. But hype about dead-ends and disasters might be driving some vendors to their knees.
Exhibit A? Well, for today it’s the latest press release from COBOL vendor Fujitsu. This week the company floated a story about success at one of its client shops, the City of West Covina. We look at the FUD language in that story and search for alternative tales of migration from vendors more friendly to your familiar technology choices. And we hear of a very large insurance vendor's migration project deadline and how it relates to disasters — of the genuine kind.
December 02, 2005
HP labs report work on 3000 updates
HP's MPE/iX engineer Jeff Vance posted a report late Wednesday night to update the 3000 community on System Improvement Ballot (SIB) upgrade projects. Yes, the 3000 engineers in HP are still working on delivering enhancements, even though the vendor will walk away from the market in less than 13 months. This is known as giving the 3000 business at HP the ending that it deserves.
Vance said his message was a quarterly update on SIB. We checked our reports to find one about 90 days earlier which covered what HP wanted to put inside the final PowerPatch for MPE/iX 6.5. (HP is still patching 6.5, as you will see below. But this fall saw the final 6.5 PowerPatch, its collection of patches in a single release tape or file.)
This time around, the news is about the march of Network Printing patches, FTP security, support for very big disks on 3000s and new CI functions. Oh, and there's a new BIND version for 3000 DNS support, and a FW-SCSI reporting tool.
You can read the full text of Vance's message posted to the OpenMPE mailing list at:
We found it interesting to compare what was proposed for lab work in March, what was ready for an update in August, and what is ready now. There's been some measurable progress, as HP reels in the sensible lines of enhancement that it cast earlier this year.
In March, Vance said HP was evaluating these projects:
1) Return the number of bytes remaining in the CI’s variable table;
2) Support a method to RESTORE files only from a specific volume set, perhaps an ONVS= parm like STORE and REPORT use;
3) Add security to FTP so that it doesn’t require Read access for the NETRC file; and
4) Support the ‘site chmod’ command in FTP.
HP also said in the spring it was figuring out if it wanted to back-port, to the 6.5 and 7.0 MPE/iX releases, these prior SIB items:
1) Network printing to non-HP printers
2) FTP server security enhancements (ftpuser, ftpaccess)
3) CI Functions (volinfo, spoolinfo, devinfo)
4) User Functions
The latest report shows that the backporting is completed for 7.0 and 6.5 network printing. 7.5 went into general release a few months back. Patches MPEMXU1B (for 7.0) and MPEMXU1C (6.5) are in beta-test status today. Vance added a note on getting this work tested — your job, as customers:
Most of these SIB patches are available on 6.5, 7.0 and 7.5, yet we still have insufficient beta testers to release the patches as General Release (GR). Many of you voted for these items so please consider being part of our beta test (BT) program.
As promised, FTP security is now available for 6.5 through 7.5 MPE/iX releases. But it's all in beta-test status. Documentation is on the HP Jazz Web site for the FTP security:
Support for disks of above 300 GB is ready for beta-testing on 7.5, and will be ready for beta not later than January for 6.5 and 7.0. MPE/iX won't be able to see any disk sectors beyond 512 GB, but commands like REPORT will now work with values greater than 24 GB. Very important if you want to attach modern, larger disks to HP 3000s in the future.
The devinfo, spoolinfo and volinfo CI functions "are close to beta-test," Vance said. These functions allow "system managers to more easily administer their system without resorting to writing more complex and privileged programs."
Bixby, who had to move away from the 3000 lab this fall after a legendary six-year run porting so much open source software to MPE/iX, has ported a new version of the BIND routines which control the 3000's Domain Name System capabilities.
Finally, HP put that SCSI tool on its Jazz downloads page, http://jazz.external.hp.com/src/
December 01, 2005
Second careerist seeks 3000 stewardship
That might be the headline of a situations-wanted advertisement for HP 3000 experts. Some have been separated from companies who mismanaged themselves into extinction, while others are no longer working because the HP 3000 they tended has been put on autopilot status, or grounded. Such is the fate of those who have good homesteading skills.
Brian Donaldson is one of these community members. He has been looking for work in the computer business for two years, after working for more than 20 among HP 3000s. Earlier this week we took note of his debut on National Public Radio, as the subject of a Hunger in America series installment. He's been contributing on the 3000-L mailing list for many years. He still has a Series 917 of his own in his Oakland-area apartment. He still has skills to offer.
Donaldson's availability is typical of the kind of opportunity intersection that this community is going to be driving through over the next few years. This week we have examined the time that it takes to migrate, based on reports from Summit credit unions. It's significant. More time than many expect, including testing. Last week we looked at Linux and open source software as a next destination for 3000 sites. Even the community's experts say that Linux has a suck-you-in factor set on high, if you have the least bit of geek in you.
Migration, training in new environments: It will all take time. Who's going to take take of the 3000's needs while your staff makes its way to the future? Perhaps people like Donaldson, who don't need a job so much as paying 3000 work. We're talking outsourcing here, a strategy that might be more attractive to management than hiring consultants.
With the amount of HP 3000 development on the wane, engineers like Donaldson might be a resource to the 3000 community that's preparing to migrate off the platform. He gave us a rundown of his recent experience on the platform:
Thanks to my Y2K bonus way back when, I was able to go buy my own 3000 -- a neat little 917LX. I am still using it almost every day, so my skills are still very much current -- COBOL, TurboIMAGE, VPlus, KSAMXL, Adager, MPEX and so on.
Having my own box has given me the freedom to write a bunch of apps that I would never have been allowed to write at any company as an employee.
1) I wanted Vesoft’s Security/3000, but they wanted way too many bucks for it, so I wrote my own. Gives me all the basics of the Vesoft product -- session logon security, console passwording, modem passwording, and an app that emulates VEAUDIT/3000 -- gives me reports of user.accts etc without passwords, PM programs, SM users and so on... If Eugene could do it, I knew I could do it too. Wasn’t that difficult, just a lot of work.
2) I also wrote a COBOL/VPlus app that gives me a list of synonyms in an Image data base. Where HOWMESSY only gives you a %age of synonyms I took my app a bit further and actually showed the offending records that are on a synonym chain -- the primary entry and the synonyms underneath it.
3) Wrote an app to monitor source code changes. Of course, any fileset can be monitored, not just source code, but I wrote it with source code in mind.
4) Inventory app for my music CD catalog. Needed some kind of indexing system but discovered that Image’s wildcarding lookup “indexing” was quite flawed. Discovered that if you wanted all the entries with the key value “FREE LIVE” the DBFIND would find it ok, but if you tried to lookup by “LIVE” the entry wasn’t found! So then I wrote my own indexing system for this app and it works great. Lookups are fast but the updating of the index keys isn’t all I wanted it to be. A bit on the slow side.
5) An app to change user.acct passwords constantly -- user defined list of user.acct passwords to be changed every “n” seconds (where “n” is user defined). Only problem with this app is that it works perfectly and changes those passwords! Quite annoyingly I might add. It changes the passwords to random values so if you don’t have the caps to do :listuser xxx;pass you will never know what your password changed to unless you go ask the sys mgr. I’ve written some other apps too but cannot remember them offhand at the moment.
While doing all this programming, I got to learn some other neat things I knew nothing about such as AIF’s, even Pascal. I got a lot of help from Stan at Allegro on the Pascal stuff. Even how to install patches, do upgrades and so on.
So I think I got my money’s worth when I paid all that money for my box way back in 1999 or 2000. I love my 3000. I told my wife Tina I would sell her before I sell my 3000. I wasn’t joking, either! :-) I’ve signed up to take an Oracle class starting January at the local community college. Not sure if it’s really worth the trouble.
If after listening to our podcast from Monday about Brian's situation, and looking over his potential, you want to make a difference in his life, contact him at BMDinSoCal@aol.com. You could save him, and others, from the cheerless alternative of Oracle training — just by setting up the interim 3000 stewardship you will need to make a migration work out.