I am not certain at what point in my life I became so detail oriented. I am one of those “a place for everything and everything in its place” people, and I rely on details in my current position as a DBA. I need details, lots of them, and I need them to be accurate in order to get to the bottom of issues. Sure, there are times when I need to weed through information, tossing out items that contradict logical thinking, but I always try to piece together the whole puzzle using everything I am given.
What I have found, however, is that people tend to believe things they hear without ever bothering to check facts. Now, I heard that Brittney had some issue the other day handing her children over to K-Fed, and to be honest I would have trouble handing my children over to him as well. But the news reports said that she was intoxicated, based upon the evidence of an ambulance coming to the home, etc., etc. Well, unless you were right there, you do not know the entire story, so you should reserve final judgement on Ms. Spears. Was she intoxicated? Yeah, probably. But could there have been another explanation? Absolutely.
Why am I telling you this? Two reasons that I call End-Arounds and Misinformation. And they are driving me crazy.
And end-around is when a person leads you in one direction only to go around the other way when you are not looking. For example, let’s say that a customer comes to you for help with some issue. You notify the customer and spend time looking into the problem (hours, perhaps), and then you see an email from a colleague saying that they will look into the issue. Why the email? Well, the customer decided that they would go to a different DBA, explain the problem, and try to get an update as to what the status is at this time. Oh, and they may or may not tell you that someone is already helping them. Well, if they wanted a status, why not ask the DBA working on the problem? Why the end-around?
Clearly, when faced with such a situation the first response from any DBA team member has to be “Let me check with so-and-so on the status, as I can maybe leverage off their research so far.” The end-around has the potential to fragment even the best of teams.
And then there is misinformation. “The server was working fine until the reboot” is a nice example. Well, it wasn’t, actually, but now that you have said that it was, everyone believes you despite all of the evidence I have to the contrary. So then we end up in a meeting, and everyone wants the answers to questions that are based upon complete misinformation. And the question are based upon rumors, and not actual facts, but you get placed on the defense. Some examples include:
“It was working before the reboot, and then it stopped. Must be a patch affecting things.” — uh, not really, no. See, no patch was applied, so that is not possible, and your stuff possibly stopped working last week, but you did not notice it until this morning.
“Can we roll back the patch” — again, no patch was applied, so no, we cannot roll it back. sorry.
“Why did it stop working” — not sure, because we have no evidence that it ever worked. Perhaps your performance just went from poor to really poor, and you only now decided to ask questions.
“Why did it stop right after the reboot?” — again, not sure, because you only noticed it the next morning, about twelve hours after the box was rebooted, so we do not have evidence that it stopped right after the reboot.
I think my point here is that we all need to make certain we only discuss the facts, and not the supposed facts. If you believe that the server stopped working after the reboot, then supply some evidence. Show me that your query ran fine one minute before the reboot, and failed one minute after the server came back online. If a patch is affecting things, fine, but make certain that a patch was applied (and make certain it was applied to the server, and not to some other server that has no relation to yours). Allowing speculation to come into play results in people having to spend time gathering evidence to refute things. And that time could be better spent trying to ascertain what the next steps should be, rather than proving what steps did not happen.
Of course, all research has it merits, and every stone needs to be overturned. And I am not saying that we are perfect and do not also jump to conclusions from time to time. All I am saying is that we all need to recognize the importance of facts versus speculation, especially when we want to have answers quickly. When somebody starts asking to have patches rolled back and it forces us to spend time proving no patches were applied, that time would be better spent working on a solution.
