[systemd-devel] systemd39: journald segfault brings down some user services
warpme at o2.pl
Thu Mar 8 13:07:37 PST 2012
On 3/6/12 2:54 AM, Lennart Poettering wrote:
> On Mon, 20.02.12 23:35, Olav Vitters (olav at vitters.nl) wrote:
>> On Thu, Feb 09, 2012 at 08:12:55PM +0100, Lennart Poettering wrote:
>>> Now, of course, the journal shouldn't crash in the first place. This bug
>>> is still something to fix, but so far nobody managed to get me a bt of
>>> this. if the journal itself crashes a coredump will be placed in
>>> /var/lib/systemd/coredump/. It would be great if somebody could generate
>>> a backtrace of that!
>> See https://bugs.mageia.org/show_bug.cgi?id=4588 and the duplicate for a
>> few different stracktraces. Mageia still has v40 atm, so every crash is
>> *very* noticeable.
> OK, so I looked and looked and looked at the code, and I think I finally
> figured it out now. Should be fixed in git, in 48496df.
> Please test!
> If I fixed the right bug then you should have encountered your problem
> only if the journal files grow quite large.
> Thanks for providing the stacktrace, much appreciated.
After 3 days of some extensive testing I have following observations:
-there is no more segfaults in journald (so far :-) ).
-As You mention, I started to have issue with quickly growing size of
journal file (after 3 day of test file had 800+ MB).
Before this change it was weeks without journal size issue - with this
change I notice quick part.filling within days.
Is quickly growing journal issue an temp.one or rather is is now by
design and users should mitigate it somehow ?
I haven't set any limits in journal.conf - so maybe I should set them.
Unfortunately there is no man for this file (or I miss something) - so I
prefer to first understand it then next modify content...
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 83 bytes
Desc: not available
More information about the systemd-devel