Depending on Other Groups
Interesting thing happened in the last 24 hrs. at work. We have a nice instrument master database - it holds all the instrument data as well as marks and such. Very nice. It's so nice that it's replicated to a read-only copy here in Chicago as well as a replicant in London for disaster/recovery purposes. It's great - except when the replication fails.
That's what happened last night. Normally, there's an explanation sent out about the "why" of why the replication failed. Changed tables, duplicate records - something caused the replication that normally works just fine to fail. But today I didn't hear what the cause of the break was. It appears to be fixed, but if we don't know why it failed, why can we assume that if it looks OK that it'll stay that way?
No reason to think so in my mind.
So when I ask the guys whose responsibility it is to maintain the database(s) "Is the read-only copy OK to use tomorrow?" and get a simple answer like "It should be good for tomorrow" I get a little nervous. What was the problem? Are we sure it's OK? I'm feeling like I'm on the beach and they're saying:
"Sure, Bob... go back in the water... pay no attention to the fins circling..."
"Oh, are we going in? No, not right now, just ate, don't want to get cramps... but you go right ahead... and splash a lot... here, hold this pork chop."
But since they are the ones responsible for the database(s), I'll listen to them and if I walk out of the surf a bloody, dismembered ghost of myself I'll say "Hey! You said it was OK!", and if there's fallout it'll be up to them, and not me, to answer those questions.
UPDATE: Well... as if I might have seen it coming, the read-only database was not ready for prime-time and I got the call at home after my server stalled on the restart due to waiting on the database. I switched it to the primary database and all was fine. Guess that means I'll have to figure out why.