[systemd-bugs] [Bug 61191] systemd-journald eats 100% CPU

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=61191

--- Comment #4 from Maxym Synytsky <synytsky at fastmail.net> ---
There were a couple of reboots since then, so things changed.
I cannot see any more EPIPE error messages. As for the rest, the issue is still
reproducible.
Currently strace shows the following:

clock_gettime(CLOCK_REALTIME, {1362639311, 615760612}) = 0
clock_gettime(CLOCK_REALTIME, {1362639311, 615791280}) = 0
writev(6, [{"<46>", 4}, {"systemd-journald", 16}, {"[21669]: ", 9}, {"Sleeping
for 17492137 ms", 24}, {"\n", 1}], 5) = 54
epoll_wait(7, {?} 0xbfcf939c, 1, 17492137) = 1
read(8, "46,5081212822,316432051764,-;sys"..., 8192) = 79
clock_gettime(CLOCK_REALTIME, {1362639311, 615929585}) = 0
clock_gettime(CLOCK_REALTIME, {1362639311, 615959991}) = 0
writev(6, [{"<46>", 4}, {"systemd-journald", 16}, {"[21669]: ", 9}, {"Sleeping
for 17492137 ms", 24}, {"\n", 1}], 5) = 54
epoll_wait(7, {?} 0xbfcf939c, 1, 17492137) = 1
read(8, "46,5081212823,316432051770,-;sys"..., 8192) = 79
clock_gettime(CLOCK_REALTIME, {1362639311, 616098321}) = 0
clock_gettime(CLOCK_REALTIME, {1362639311, 616128761}) = 0
writev(6, [{"<46>", 4}, {"systemd-journald", 16}, {"[21669]: ", 9}, {"Sleeping
for 17492137 ms", 24}, {"\n", 1}], 5) = 54
epoll_wait(7, {?} 0xbfcf939c, 1, 17492137) = 1
read(8, "46,5081212824,316432051776,-;sys"..., 8192) = 79
clock_gettime(CLOCK_REALTIME, {1362639311, 616266463}) = 0
clock_gettime(CLOCK_REALTIME, {1362639311, 616296821}) = 0
writev(6, [{"<46>", 4}, {"systemd-journald", 16}, {"[21669]: ", 9}, {"Sleeping
for 17492137 ms", 24}, {"\n", 1}], 5) = 54
epoll_wait(7, {?} 0xbfcf939c, 1, 17492137) = 1
read(8, "46,5081212825,316432051782,-;sys"..., 8192) = 79
clock_gettime(CLOCK_REALTIME, {1362639311, 616434289}) = 0
clock_gettime(CLOCK_REALTIME, {1362639311, 616464762}) = 0
writev(6, [{"<46>", 4}, {"systemd-journald", 16}, {"[21669]: ", 9}, {"Sleeping
for 17492137 ms", 24}, {"\n", 1}], 5) = 54
epoll_wait(7, {?} 0xbfcf939c, 1, 17492137) = 1
read(8, "46,5081212826,316432051788,-;sys"..., 8192) = 79

Descriptors 6, 7 and 8 are as follows:

# ls -l /proc/21669/fd/6
l-wx------ 1 root root 64 Mar  7 00:29 /proc/21669/fd/6 -> /dev/kmsg
# ls -l /proc/21669/fd/7
lrwx------ 1 root root 64 Mar  7 00:29 /proc/21669/fd/7 ->
anon_inode:[eventpoll]
# ls -l /proc/21669/fd/8
lrwx------ 1 root root 64 Mar  7 00:29 /proc/21669/fd/8 -> /dev/kmsg
#

BTW, when I set Storage=none the issue is not reproducible so I think this
might be something with writing the logs.
So at the moment I can control and trigger this behaviour - set Storage back to
the default value and restart journald.
If you need strace just after the journald restart I can try to get it.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/systemd-bugs/attachments/20130307/0c2c8688/attachment.html>