First off, the swimming update. I got 2000 yards today, bringing my total to 55,000 yards and leaving 1,705,000 to go.
Secondly, I’ll have to vent a little. Since last Friday, we’ve had a problem with one of the main programs we use at work. The result has been that we’ve been working at less than optimal capacity, although still functioning. During this time, I called the support folks who created the program about 15 times, and did not ever get a live person. Even worse, after the first robot, there wasn’t even an automated voice to complain to.
In addition to those multiple phone calls, I sent another dozen or so e-mails back and forth with one of the guys there. (BTW: he gave me his direct line, and I called it three times with no answer too.) Since they’re located two time zones away, he had me set up a program where he could remote access our server. He did that yesterday, said he thought he found the solution but wanted to test it on their virtual machine on his end first and asked for me to set up a meeting today. I did so. He poked around in our server today, and since I could watch what he was doing it quickly became apparent that he had no idea what that was. Indeed, his “solution” obviously…wasn’t.
I finally decided to go to lunch after an hour of watching him intermittently do stupid things with extended periods of him doing nothing intermixed. When I got back to the building with my lunch, turns out that the power in the building had gone off. It came on about an hour later and I had an email from this guy asking when I could set the connection back up. I told him what had happened and said I could do it then. No response.
So I finally went down to our MIS department and asked one of my friends there (the same one who’s gotten me hooked on Eve-Online, by the way, and he’s our local server guru) if he had any ideas about a different issue that was happening that I thought might be tangentially related to this current problem. Turns out that it wasn’t related, but in the process we began to poke around a few files, found a corrupted database, and he restored it. I went back to test our server and it seemed to start fine but then crashed, re-corrupting the database. So we got an older version of the file (by one day), then I got my former supervisor (who is now merely a co-worker) and said, “I don’t think this database has anything to do with what we’re actively using; it looks to me like a legacy database. Would you agree?” (Note: I didn’t put it in those words since he’s not a techie, but I’m also not at liberty to say exactly what I said due to work confidentiality, but this gets the gist across.) He did. So I said, “I think we’re fine if we go with a really old version of the database if we have to.” In the end, we didn’t—we just used the one-day-old version, but it’s still thrilling knowing that your problem is caused by something you don’t even use anymore.
In any case, at this point I had already figured out exactly what was wrong. Last week, one of our new people had problems with the software, and it turns out it was exactly the same problem that a person two years ago had when she was also a new hire. I was able to identify what the problem was two years ago and our help desk fixed it, but I couldn’t remember what the solution was since it had been two years. Thankfully our help desk logs all those calls, and they pulled up what they had done and fixed the new person’s computer. After that fix, these current problems started, and I theorized that one of the batches our new person had been working on before that computer was fixed was itself corrupt and it was causing the corruption of the rest of the database.
Since we knew which batches had been keyed by the new person on that day, we excluded them from the process the server was supposed to do; it fixed the problem with the server corrupting the database. Then, with backup copies of the database in hand so that we wouldn’t have to keep transferring them from the back-up tape, I worked through until I had found the specific batch causing it to crash and corrupt everything. We can’t repair that batch, but I even found a work-around that will leave us fully functioning with no adverse side effects.
It took just over an hour for me, with the help of two other coworkers, to do what our support couldn’t begin to touch in 15 phone calls, a dozen e-mails, and a wasted week. Yet I still don’t get paid techie money, and somehow I have a feeling we’re going to get billed from the support people even though we solved the problem for them and I personally think that means they should pay my paycheck this week.
So harrumph.
