I feel like this is probably something I should be able to find info
on in an existing faq or archive, but I haven't been able to find
anything. Are these daemons supposed to use a whole core 24x7? The
setup *seems* to be functioning normally in initial testing
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ Flags
WCHAN COMMAND
4576 root 20 0 580m 468m 4528 S 100.0 2.9 1259:56 ..4.214.
futex_wai cmds
4480 root 20 0 138m 39m 2904 S 100.1 0.2 1259:39 ..4.214.
futex_wai cmon
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ WCHAN
Flags COMMAND
2540 root 20 0 1140m 861m 3656 S 100.0 2.7 1269:59
futex_wai ..4.214. cosd
As you can see almost all of the time is spent in futex wait, strace
reveals a billion polling events on one of the threads (on each of
cmds,cmon,cosd), e.g:
[pid 4583] poll([{fd=7, events=POLLIN|POLLRDBAND}, {fd=5,
events=POLLIN|POLLRDBAND}], 2, 0) = 0 (Timeout)
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
