Actions
Bug #55669
openosd: add log for pg peering and activiting complete
Status:
New
Priority:
Normal
Assignee:
-
Category:
-
Target version:
-
% Done:
0%
Source:
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
rados
Component(RADOS):
OSD
Pull request ID:
Crash signature (v1):
Crash signature (v2):
Description
osd: add log for pg peering and activiting complete
This log is added to PeeringState::Active::react(const AllReplicasActivated &evt)
to pair with the PeeringState::start_peering_interval() log,
So that under the default log level of debug_osd(1/5),
the completion of each pg peering can be traced back through ceph-osd.x.log
Because under some exceptions, we need to know clearly
whether pg has completed peering at the time when the exception occurs,
and whether it can handle client IO.
Signed-off-by: Jianwei Zhang <jianwei1216@qq.com>
PeeringState::start_peering_interval() ///This is the log of the original pg peering start
{
psdout(1) << __func__ << " up " << oldup << " -> " << up
<< ", acting " << oldacting << " -> " << acting
<< ", acting_primary " << old_acting_primary << " -> "
<< new_acting_primary
<< ", up_primary " << old_up_primary << " -> " << new_up_primary
<< ", role " << oldrole << " -> " << role
<< ", features acting " << acting_features
<< " upacting " << upacting_features
<< dendl;
}
PeeringState::Active::react(const AllReplicasActivated &evt) ///This is the log of the newly added pg peering->activating->active end
{
psdout(1) << __func__ << " AllReplicasActivated : Peering->Activating->Active over"
<< " PG_STATE_ACTIVE=" << ps->state_test(PG_STATE_ACTIVE)
<< " PG_STATE_WAIT=" << ps->state_test(PG_STATE_WAIT)
<< " PG_STATE_PEERED=" << ps->state_test(PG_STATE_PEERED)
<< " last_epoch_started=" << ps->info.last_epoch_started
<<dendl;
}
The purpose of adding this part of the log: it is convenient to locate the log time from the start to the end of a single pg peering
Updated by jianwei zhang almost 2 years ago
Actions