I have a second blog in which I don’t go on any rants etc, just post technical stuff. I just posted a entry today on Linux memory management, especially with databases. One thing that was confusing to me originally is that, for instance, on an Oracle server, I have a free output like this:
total used free shared buffers cached
Mem: 16027 15889 138 0 48 14673
-/+ buffers/cache: 1166 14860
Swap: 5122 486 4636
So there is like 15 gigs cached out of 16 gigs total. Obviously with 2 million oracle processes, it is hard to tell how much memory it is actually using, and I stumbled on the ‘ipcs’ command which lets me see those shared buffer sizes.
If you have a bit more insight into this and can proof read the blog post, that would be great to let me know if I made a mistake. Thanks!
Related posts: