Troubleshooting

In Anthropology we learned that stresses on one part of a system oftentimes had unexpected expressions in an entirely different unrelated area of that same system. In IT that lesson shines through regularly, and is one of the key lessons that I have taken to heart and use as a prime example of why my degree in Anthropology did not go to waste.

Today my app went down. Well...it seemed like it went down. The end user thought it did at any rate - because the end user could not get into my app. That would be one definition of "going down" in a software sense.

However, upon closer examination, the app was just fine - it was happily sitting in it's application server waiting futilely for connections that never came...and the connections that it did have were suddenly bereft of user information, tossed around in a sea of namelessness trying aimlessly to connect with nothing.

Aha. It wasn't actually the app. Information that was supposed to have been passed to the app wasn't being passed, and I wasn't doing my job of properly notifying people to this effect. So I worked on fixing notifications and in the meantime information started flowing...and things looked to be hunky dory.

Later this evening another issue cropped up - seemingly exactly the same - but then under the surface it was completely unrelated and had different symptoms.

Or did it? I took most of the same steps, but now I had more information to go on, and followed a different path to troubleshoot the issue. I found a different issue this evening. Still external to the app itself. But dependent on the type of information that the app provides.

Ah, but one could not have exactly forseen the types of issues that would have required the logging statements written today. So were they truly unhelpful messages? Or were they simply young and naive messages that have now been given a chance to blossom and mature just a little more?