Steve Shockley wrote: > On 3/5/2010 7:42 AM, Nick Holland wrote: >> And yes, this is just the tip of the iceberg with vmware quality >> issues, but that one was really, really easy to understand. > > So, you're saying VMware *is* "enterprise-ready", then? Like Blackberry > Enterprise Server, CA Message Manager, or any number of products that > appear in a "quadrant".
yeah, of course. Based on the products that stick "Enterprise" in their names, the computer world definition appears to be something like: * High priced ("Our product will save you X dollars a year! We'll be charging you 0.5*X dollars a year, but you will be way ahead! It's worth it!" Of course, the X saved is adjustable and impossible to really measure, the 0.5*X is a fixed cost you /will/ be paying). * high down time (caused by the rest of the things on this list) * arcane knowledge required (I don't mean "expert" knowledge -- you have to understand a lot of things to properly manage an e-mail system, but most "enterprise" products seem to deliberately ship in crippled mode and require FIXING to bring up to usable levels, using all kinds of voodoo syntax *cough*Cisco*cough*) * Insecure by design * really cool blinky charts and graphs to show the managers who have no f'ing clue what they mean, but they look good. * good to put on your resume, because it is far more important to be "Enterprise grade" than to do good work(*). * many cases of Munchausen by Proxy Syndrome (i.e., administrators "selflessly" giving up their weekend and evenings to come and rescue the company from problems that never should have existed in the first place). I've seen cases where it isn't just the administrator's fault -- I've seen managers who were completely unimpressed by people who's systems Just Worked because they were did Good Work, and promote those that "Work really hard for the company" by bandaging crap that should have been properly FIXED, or whined that the systems were not "Enterprise Grade". Yes, I watched a really good administrator get passed over for for promotion because his reports basically said, "all is good" and things Just Worked, but someone else of considerably less skill was saying, "oh, it is all junk!" (*) Good work: So far, I don't think this phrase has been co-opted and corrupted by any crapware maker, so let's claim and define it: * True value price, not "ability to pay". * Return on investment not requiring hallucinogens to see. "They say we'll save $10 million a year!" "Dude, our total cash flow is only half a million". My favorite example: someone told the owner of a company that 20% of the employees were file clerks, and that the total payroll was $30M/yr. Thus, if they could get rid of the [low paid] file clerks, they could save $6M/yr [yes, that was the argument: file clerks and VPs all earn the same amount of money by this logic], and a document imaging system would pay for itself in a year! This set a chain of events in motion which has had..uh..interesting effects on the company, and as far as I know, not a single file clerk has been laid off. The top people in the IT department are all elsewhere, though (and no, I'd never believe this story, other than the details were provided to me by someone I trust who got it from someone who would know...and that I've seen evidence the owner had some serious problems with reality). * Minimal total down time. Infrequent, short is ok and expected. infrequent, long is not. Frequent, long is right out. * True redundancy and ability to withstand a realistic fault in an acceptable way. * Simple enough to avoid creating bizarre faults that can't be handled in an acceptable way (i.e., simple systems have simple problems. Complex systems will have things happen you can't imagine and didn't plan for). * Expectation that things won't go well at some point and plans in place to deal with it when it does go bad. * Provisions to move to a new solution when better solutions become available and to replace existing hardware when it is old and beyond its serviceable life (how many projects have you seen which lingered on and on on old crap hardware because no one could figure out how to re-implement on something modern?) * Acceptance that no one in your company has ever done EXACTLY this project before, and there will be learning taking place. The only people who have precisely "done that many times" are either liars or total screwups, usually both. * Spending a small amount of money on properly run pilot projects before spending large amounts of money on the final project is money well spent. I should probably clean up and refine these lists, that was just a quick toss-together. yeah, I'm venting, but I'm increasingly finding this industry embarrassing to work in. Nick.