Person: Is there a problem this morning?
DBA: A problem with what?
Person: Application “X” is throwing out errors. Do you know what’s going on?
DBA: That’s wierd, as far as I know nothing has changed on that project recently.
Person: They put some new software live yesterday, but they are sure that’s not the problem!
DBA: Which server are they seeing the problem on?
Person: Server “Y”.
DBA: The applications on that server don’t connect directly to the database. It doesn’t even have an Oracle client loaded.
So let me see:
- The specific application does not, and has never, connected to the production database.
- Some new software went live yesterday.
- The problems they are seeing must be related to the production database. It can’t possibly be due to the new software put live yesterday!
I’m sure there is logic somewhere there, but I’m just too stupid to see it 🙂
Cheers
Tim…
Fret not. I’m confused as well.
(*_*)
Person: They put some new software live yesterday, but they are sure that’s not the problem!
Obviously a work of fiction. In real-life(TM) that line would read:
Person: They swear they have changed nothing.
Are you sure you didn’t read my mind when writing this post? We see it many time …
The problems is always with (pick one or more)
* oracle
* 9i
* RAC
* CBO
* CBO
* CBO
and is never with the code or the design itself.
I also believe “no one changed anything” is a private synonym for “I am clueless but I am not the problem”.
RJamya
When performance/whatever goes down the tubes, we all know “it is ALWAYS the database, ALWAYS”
They know it works, because they saw the demo of it by the salespeople and that went off without a hitch. As long as the salespeople didn’t sell them a bum product, it should be fine to put it live into production, right?
———————————–
Reminds me of a problem I ran into yesterday. I wanted some soda, but didn’t have enough change on me to get some. Walked out to my car to get some change and then got my soda from the vending machine. I get back to my desk and one of my coworkers had left me a few im’s that said “Stupid constraints.”, “Oh wait, nm, problem solved”
Curious, I asked him what the problem was. He had switched two sequence inputs to a procedure and the constraint was saying that there were no referential matches for those values. Hooray for constraints preserving integrity!
Just another example of how the database is blamed when it is reflecting a user error. There’s a great saying for it: PEBKAC (Problem Exists Between Keyboard And Chair) that is a rather easing way to admit error. “Oh, it looks like another PEBKAC error, I’ll go and fix it.” is another way of saying “Oops, I screwed up.” 🙂