On Fri, Oct 07, 2011 at 02:49:22PM +0530, Ritesh Raj Sarraf wrote:
> This one might be a difficult one to notice. But if you run ulatencyd on
> a machine that you don't often reboot, you might be able to reproduce my
> bug.
>
> On my laptop that doesn't reboot frequently, over time, ulatencyd daemon
> dies. I had to run it in debug mode with hopes that some extra
> information could be extracted. Unfortunately, not much.
>
> Following is the last set of messages logged in by the daemon.
>
> 14:37:13 rrs@champaran:~/Community/Packaging/pkg-iscsi/targetcli
> (master)$ tail /var/log/ulatencyd.1
> DEBUG: run filter: /etc/ulatencyd/rules/lxde.lua: LxdeRunnerFix
> DEBUG: run filter: /etc/ulatencyd/rules/system.lua: SystemIdle
> DEBUG: run filter: /etc/ulatencyd/rules/system.lua: SystemBg
> DEBUG: run filter: /etc/ulatencyd/rules/zz_fixers.lua: MediaIO
> DEBUG: took 0.00. schedule:
> DEBUG: scheduler filter:true
> DEBUG: took 0.00. complete run 5459 took 0.04
> DEBUG: IO: set group isolation on dev dm-0 to 0
> DEBUG: IO: set group isolation on dev sda to 0
> DEBUG: cleanup daemon
> ^^^^^^^^^^^^^^^^^^^^^^
This is just an hopeless try but, ulatencyd gets restarted daily by
logrotate and given #643582 it may be the cause of ulatencyd stopping.
This should leave some traces in the log (which is not your case) but, in
the absence of alternative solutions as of now (and now that the above bug
has been fixed), it may be worth checking just to make sure. It doesn't
explain the "sh" mystery though.
Cheers
--
perl -E'$_=q;$/= @{[@_]};and s;\S+;<inidehG ordnasselA>;eg;say~~reverse'
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
