On Wed, Oct 12, 2016 at 2:03 PM, Piotr Kliczewski <pklic...@redhat.com>
> What I meant is that you sent logs created at different times so I was
> unable to match both ends (engine and vdsm.
> It would be great if you could sent the logs from the same time.
Probably there is something I don't understand...
I collect logs at 11:05
I copy both current engine.log and current vdsmd.log and update them to
Due to how ovirt-engine and vdsmd are logging and to how they are rotating
1) the current engine.log initial and final lines at that time are
2016-09-27 03:16:08,460 INFO
(DefaultQuartzScheduler1) [1e566c35] Fetched 5 VMs from VDS
2016-10-12 10:51:36,336 INFO
(org.ovirt.thread.pool-8-thread-3) [da0fb3b] Finished automatic refresh
process for 'ISO' file type with success, for storage domain id
quite obviously no more logging during latest 14 minutes.... and so no more
lines inside the file...
2) the current vdsmd.log initial and final lines are instead
r/bin/taskset --cpu-list 0-31 dd
flag=direct,fullblock count=1 bs=1024000 (cwd None)
SUCCESS: <err> = '1+0 records in\n1+0 records out\n1024000 bytes (1.0 MB)
copied, 0.00455719 s, 225 MB/s\n'; <rc> = 0
tipically vdsmd.log rotation happens every 1 hour and in fact that
particular vdsmd is now named (at 14:43) vdsm.log.3.xz and in total
contains from the already shown line
/usr/bin/taskset --cpu-list 0-31 dd
iflag=direct,fullblock count=1 bs=1024000 (cwd None)
handler for 127.0.0.1:47818 stopped
when it has been rotated....
Please explain what is wrong with my approach giving you the information
needed.... Do you need more vdsmd.log files in the past?
Users mailing list