At work I spent most of the day deleting old tables from my Oracle database. The other name for that is maintenance, or just paying down some technical debt.
One of the things I learned about my boss after working with him for a couple of years (yeah, I'm a bit slow sometimes) is that I should never ask him if he wants to keep an old version of something. If he had his way, our website would have every version of every calculation we've ever done on it.
And he would only ever look at the most recent one.
In with the new, out with the old
The problem with his approach is that if you ever have to change one thing, now there are fifteen things that might be changed along with it. Are any of them still in use? I mean, something is referencing them. But is it important enough to spend two weeks trying to find another source for the data, or am I going to spend all of that time and then realize that it's all for something that hasn't been looked at in two years?
So, yeah. Now I don't give him a choice. I try to clear away old stuff as I can, though that doesn't always happen.
Data hoarding
Today I deleted queries and tables like:
- Data that is six years old. It might be useful one month back, but six years? Nope.
- A project that I put together five years ago. We used it for a couple of months, and then everyone (even me) forgot about it. I only have a vague memory of what the original purpose was.
- Data for a technology that is no longer in existence in the USA. (It's probably not used anywhere in the world, but I don't know that for sure.)
- Tables created by people who left the company three years ago.
- Tables for projects that were originally in our group, but then moved to a different group (and thus a different database), and have since moved again.
Data Recycling
Ripley in fancy dress |
No comments:
Post a Comment