Wednesday, 12 December 2007

Mushrooms and those that don't learn...

SO we had this "fiasco" when a server got ordered, the guy though that it'll come with the necessary drive-brackets (surprise!!!!) to be able to buy the el-cheapo drives from the cafe around the corner as the true black-and-brown drives are 2 to 3 times the price... ends up to buy the smallest drives from black-and-brown to get the drive-brackets to be able to use the el-cheapo big-drives... still a "bargain"... but that's water under the bridge.

Then the request for backups (again after a near crisis situation is threatening the business's livelyhood) they decided it's time for the much taunted backup solution. No problem there either.
It's only that you first have to show them the light (and please, don't open the curtains, just give them a flash light --- not the LED version, but the nearly run down filament type--- to get their eyes acclimatized) about what is a proper backup solution given the situation.

Then they order it. (Again I've been the mushroom to find out about it after I've asked about it)

Then they mention it's a type-X controller. You tell them it should've been a type-E controller... "No, but the documentation [they] I could find, and the response from Black-and-Brown's guys is that the type-X will fit this server" ... yeah right.

/me goes to net: Gargle for Black-and-Brown Big-Server-designation [Enter]

/me finds "official" website from Black-and-Brown: Type-E controller (There are something that looks like an x in the name... perhaps thats the problem...). "official" other documentation about what type of controllers that'll do backups to the proper backup solution needed: Controllers with Type-E connector.

They've ordered the Type-X connector because: lead time is shorter....

They don't learn, that these things needs to be planned up front.

They don't learn that you ask those that have been swimming in that pound how those sharks bite.


IT uptime and Directors

So the client wants to move to another/secondary/extra/redundant connectivity provider.
No problem, I actually advised and requested them to do this many months ago as the current provider is iffy at times...

Even more so they want to load another "optimized" (or something) piece of code in the place of the current running code. No problem either with that, as I actually would advise in that direction for several reasons, however the client needs to understand a few "changes" etc. which they've accepted. No problem in that either.

Then the client wants to remove a server from the stack so that the second provider can load their software on it. I warned and resisted it for stability reasons. Still the Director was insistent that they need it sooner for this that and these reasons. Several of them (mostly all of them) I have no problems with, however I advised on a different road (that some how, because I'm the mushroom, didn't get a cost value to me) that the director was not prepared to take. Time etc.

So I implemented the change. And as expected (do I have to mention that I hate it to be right???) things break, they can't do their sales etc. and the Director comes down on me like a ton of bricks.... things about warnings etc. etc. etc. and then ordered me to put the server back into production immediately...