I’m on my way to Dallas for the HTG Summit. Got a ton of shirts and associated swag, stop by, say hi, bitch about my products and I’ll fix em. The usual.
OWN had the pleasure of resurrecting the hot plate blade center back to life overnight. Had we not gotten this together we would have looked like such losers I’d be afraid to even walk around Dallas.
Umm, yeah, I couldn’t keep a 4-blade $20K blade center together. My bad. But please give me your colo business!
The rough thing is, there is no way to win any sympathy points at all with your peers. But here is the messed up part. There was an extra blade in the blade center that NOBODY can account for. It’s not ours. It’s not the developers. It wasn’t on the purchase order. Nobody knows where it came from, we just know that the extra ram that was supposed to go into isn’t compatible with that motherboard. Here is where insult turns to injury: In the system that the system does fit and boots with full compatibility and speed the heat output is so high that it triggers the heat sensor on the CPU and marks the system down for overheating, shutting it down. And just when you think you’ve been kicked enough, and you tell the monkeys you’ll hang them upside down about not being able to find a server in 20” of vertical space – you notice that the overheating server blew off the label from the blade.
Vlad: It was a hardware failure.
Dave: Should have gotten a Mac. OS X never overheats.
Vlad: One node failed.
Erick: Have you heard of this new technology called clustering?
Vlad: But I didn’t set it up
Karl: Oh, tell me a tale of how you can’t keep $20,000 worth of server alive!
Vlad: But it was the new third party RAM that we didn’t order that overheated it.
Mark: You know they have heatsinks for RAM now, it’s new.. maybe 6-7 years old?
No matter which way you spin it there is no running away from five nines. There is just no way to gain any sympathy when you spend so much $$ on the stuff.