So I was wondering around the web, and cant seem to find any decent Treads trouble shooting guide. Ans at the same time my server spiked yo 3k threads. So maybe someone could give me some tips, or possible links where I could solve my problem.
I am running standart LAMP server.
Ok, so I got the spike of
3.00k threads from my munin status. At the same time the
processes spiked up to a 2.2k. The
average nr. of threads is 1.5k. The
average nr. of sleeping processes is 1.3k. As well as that I check that I hade spikes in Swap, RAM-committed, ans MySQL Threads up to 700 threads ! Ok, so it probably means that some programmer did something interesting.
# ps aux | wc -l 1424 ps aux | grep apache | wc -l 974 ps aux | grep mysql | wc -l 3 ps aux | grep root | wc -l 475
Now I understand the apache processes, I know i have to optimize it. But the root processes it something that really disturbs me as I am not sure what those processes are and google does not to really specify them either. So i did
ps aux | grep root
I got load of different processes:
[xfsconvertd/19] [xfsdatad/19] [xfslogd/19] [ext4-dio-unwrit] [kthrotld/15] ... etc ...
The thing thats weird is that most of them are sleeping:
ps aux | grep S | wc -l 1414
Now should I worry ? and maybe I should kill the sleeping processes ? If so, which one ? Or what steps should I take from stopping it happening again ?