RAC Consulting Sponsor Message

August 2001

You’re Right, I’m Left, They’re Gone

By Scott Hirsh

By definition, we in e3000 land didn’t “get it.” We stuck with tried and true technology that evolved prudently with the times, avoiding the latest flavors-of-the-month. So as the clock strikes midnight on the new economy, turning it back into the pumpkin of the old economy, a few of our vendors or business partners may have disappeared into the void. Depending on how heavily we integrated these vendors into our environments, this can either be viewed as sweet justice or “sweet Jesus, what now?”

Always Have a Plan B

One of the first lessons taught when I became a VISTA volunteer was “always have a plan B.” If you were a Boy Scout, it was “always be prepared.” Wherever you learned your survival skills, the lesson could not be more pertinent. As someone who has consistently advocated the use of third-party tools due to the advantages of vendor support, the loss of a vendor can pose a big problem.

For example, there is a product sometimes known as the “Swiss Army Knife of e3000 utilities.” This tool is so versatile that it can be deployed to perform the functions of almost any utility on the market: job scheduling, file maintenance, log scanning… ad infinitum. Relax, those guys are not going anywhere. But if they were, many shops would be in deep kimchee because this tool gets embedded in so many job streams, command files and UDCs. So you get the idea: If one becomes heavily dependent on a vendor and the vendor goes away, one must deal with conversions or possibly living without the functionality of those products.

If you’ve been around long enough, you’ve seen ’em come and you’ve seen ’em go – even before there was a new economy. So as everything old is new again, and to reinforce what you already know, some worst practices to ponder:

Worst Practice 1: Not Having an Exit Strategy

Let’s take a simple case based on one of my core areas: job scheduling. Job schedulers are marketed as “plumbing.” That is, they’re low enough in IT infrastructure that ripping it out and replacing it is a miserable exercise. And yet, there’s been a fair amount of shakeup in all areas of third-party system management tools, job schedulers included. What’s particularly nasty about job schedulers is that scheduler-specific variables can be embedded in job streams to make jobs generic and, hence, fewer and easier to maintain. However, if the job scheduler and/or company go bye-bye, you are left with a lot of loose ends to deal with.

So what you want is a way to get the most out of the product without locking yourself into a vendor. In the case of a job scheduler, a good way to trap yourself is to embed the vendor’s specific code into your JCL. That way, the amount and variety of code are maximized, making both identification and eradication most difficult. Instead, you can imbed a short variable that represents longer code strings. That way, there will be a lot of overlap, fewer strings to search for and easier eradication. Additionally, an alternate vendor has a better chance of helping you convert from your current scheduler to the new scheduler.

The key here is always to implement with at least consideration for the day when you might need to rip out the existing product and convert to something else. I know, easier said than done. But not only does having an exit strategy position you in case your vendor disappears, it also gives you leverage in license negotiations, as vendor reps are adept at smelling the blood of the trapped.

Worst Practice 2: Not Paying Attention

The technology treadmill turns faster every day. It’s hard enough to keep up with multiple products on multiple operating systems and network protocols. And let’s not even get started on security. So it’s perfectly understandable if you didn’t notice the warning signs of vendor failure. Okay, so you haven’t had a new version since Clinton’s first term. Or the company hasn’t been seen at HP World since it was called the Interex Conference. So what, they only charge $300 a year maintenance, and their product works just fine in “compatibility mode” (whatever that is).

But all it takes is one significant – to you – vendor disappearing act and you’ll pay more attention next time. I’ve had cases where I didn’t know a vendor was gone until I got a mailing from another company offering me “competitive upgrades.” A better way to live, if you have the bandwidth, is monitoring the 3000-L mailing list, this very Newswire, user group meetings and the usual industry rags. The Web, of course, is a godsend in this respect. As we’ve said before, in our line of business anything exciting… is bad!

Worst Practice 3: Overreacting

Okay, some software contains a time bomb. But not all, and if you’re big enough you can negotiate out the time bomb. Assuming the product won’t self-destruct, guess what? You have time to deal with your vendor’s demise — unless you’re changing your environment in such a way as to affect the software (OS update, box swap, etc.) The software doesn’t know whether the vendor is alive or not, so don’t panic when you get news of a merger (most typical) or something worse.

Sure, holding up a planned OS update because your critical utilities will stop working is a bit of the tail wagging the dog. However, one shouldn’t automatically assume that a product will stop working when the next OS version is installed. (After all, that kind of compatibility is one of the strengths of the e3000!) We all perform due diligence and call all vendors to ensure that we have a supported product version for the new OS. But worst case, you can run the software on a test box with the new OS and see for yourself (a la Y2K). Also, a trial balloon via 3000-L works wonders. Not to mention, you won’t be the first one on the new OS, will you? Now that’s a real worst practice!

Worst Practice 4: Bad Choices

As much as we like to root for the underdog, in business we must play the odds. Unless there is a compelling reason to buy a critical piece of system software from the Okee-Dokee software company – and there often is – it is prudent to stick to the big boys for anything critical. And by big boys, I mean either the obvious (HP) or the almost as obvious, those who have been around many years and have high customer satisfaction (Adager, Robelle, VESoft, etc.) Should you decide to give the latest upstart your business, be honest about the risk you are taking. As we know, many of our favorite vendors were once young upstarts. But so were many that are now roadkill or fond memories.

Worst Practice 5: Suffering in Silence

A lot of bad behavior is encouraged when users suffer in silence. Often, the early warning of trouble is when a vendor starts asking for money a little earlier than expected. Or the sales reps are a bit too anxious for comfort. When you begin to suspect that something is awry, it’s time to start inquiring among other System Managers.

We’re not talking about defamation here, or yelling “fire” in the auditorium. But either subtle or overt warning signs should be discussed among peers. I can personally attest to the value I’ve received from others who have shared this kind of information, and witnessed what happens when you charge into a deal blind.

For example, a previous employer purchased an accounting package (not on the e3000, alas) on New Year’s Eve, because the deal was good for “today and today only.” Little did the CFO know that (a) the package hadn’t been tested on our combination of hardware, OS and RDBMS (and didn’t work); (b) there were almost no customers of this product; and (c) it would be discontinued in less than one year. If the CFO had sought out more information about this package – or later shared her experience – she wouldn’t have made a big mistake, and neither would the people who bought the software weeks later, just before it was discontinued.

Reality Check

There are some who would say that the e3000 is a risky place to be these days, that HP’s commitment to the platform puts it in the high-risk category of the above discussion. Given the recent turn of events for the likes of Sun, Compaq and several Linux vendors, it’s hard to say what’s risky and what isn’t. And that’s the point here: Nobody can predict the future.

So to be honest, we’ve all seen favorite vendors come and go through the natural cycles of the marketplace. But there was a time when we didn’t factor in the possibility of a vendor disappearing, and I, like many of you, paid the price in conversions at gunpoint. We won’t always bet on the right vendor to take us the distance, but we can at least be prepared for all the contingencies.

Scott Hirsh (scott@acellc.com) former chairman of the SYSMAN Special Interest Group, is an HP Certified HP 3000 System Manager and founder of Automated Computing Environments (www.acellc.com, 925.962.0346), HP-certified OpenView Consultants who consult on OpenView, Maestro, Sys*Admiral and other general HP e3000 and HP 9000 automation and administration practices.