This afternoon, we resized one of our servers from a 4GB linode to 12GB. Immediately following that reboot, I noticed that memory usage for buffers was *insanely* high - higher than I've ever seen on any machine I've touched. Here is the a munin memory graph with data pre and post migration:



Data that munin is displaying is corroborated by the output of "free":

[user@host ~]$ free -m
             total       used       free     shared    buffers     cached
Mem:         11967       9371       2595          0       7367        542
-/+ buffers/cache:       1462      10504
Swap:          255          0        255

Now, I'm well aware of the kernel's usage of unused memory for cache, but my understanding of buffers is that buffers are used to temporarily store writes until they've been committed to disk. Is that a correct understanding?

Now, three things (that I know of) changed during the migration.

1) We picked up the extra 4 cores that were given out earlier this week.
2) We're on the "Latest 64 bit" kernel, which is now 3.7.10-x86_64-linode30. Previously we were on 3.0.18 I believe.
3) We went from 4GB RAM to 12GB.

So...any thoughts about this? I can't make heads nor tails of it. This is a webserver, with very little disk IO (database is hosted elsewhere). I could see high buffer usage on a server with very high write IO, but I can't explain this in our case.

_________________

I think buffers are used also (among other things) to hold network packets that arrived and are pending processing, too. How's your network graph looking?

_________________
rsk, providing useless advice on the Internet since 2005.

Good question.

Network traffic hasn't changed significantly:

_________________

try downloadning meminfo.py, i don't think it's packaged.
I'm not an expert on its output, and I think it might not show how memory is used by the kernel, but, you know, it's a good tool for those that need more details into memory usage.