the corner office

The first of this month marked 1 year since I signed up and rented this virtual server. It's been rather well-behaved:

colin@colinpretorius:~$ uptime
 23:31:08 up 199 days, 17:51,  2 users,  load average: 0.00, 0.00, 0.00

I have no idea why it got bounced. Wasn't me, promise.

I also noticed that the kind folks at RoseHosting have bumped me up from 256MB to 512MB of memory somewhere along the line.

colin@colinpretorius:~$ free -m
             total       used     free  shared  buffers  cached
Mem:           512         30      481      0        0       0
-/+ buffers/cache:         30      481
Swap:         4000        611     3388

I've been busy on the server this evening, which has kicked up some dust. The server's been averaging around 25 megs of in-use memory for a while now. The swap file's rather big, but if the OS thinks the gumpf can be paged, it can't be that important. Who says Java apps need to be expensive? I could probably store the whole damned blog in memory, and get away with it.

Anyhoo, the fact that the server's been going for a year, meant I had some emergency admin to do. A year ago I decided I didn't like the way logrotate handled my log files, and found a utility called cronolog which suited my needs far better. Cronolog wasn't an active project though, and I wasn't in any rush to make the change because logrotate would only start dropping log files after 52 weeks.

A week or so back I realised that a year was already up, and it was about time I tweak logrotate or swap to cronolog. Needless to say, I could remember very little about how to do either, so I did some homework, and tonight took the plunge and switched to cronolog, which once again seems to be active. I wrote some notes to remind myself of what I did and how I did it.