Bug 1188451 - sosreport should collect /var/log/glusterfs/geo-replication* logs
Summary: sosreport should collect /var/log/glusterfs/geo-replication* logs
Keywords:
Status: CLOSED EOL
Alias: None
Product: Red Hat Gluster Storage
Classification: Red Hat Storage
Component: sos
Version: 2.1
Hardware: Unspecified
OS: Unspecified
high
medium
Target Milestone: ---
: ---
Assignee: Raghavendra Bhat
QA Contact: storage-qa-internal@redhat.com
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2015-02-02 23:49 UTC by Dustin Black
Modified: 2015-12-03 17:21 UTC (History)
11 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2015-12-03 17:21:03 UTC
Embargoed:


Attachments (Terms of Use)

Description Dustin Black 2015-02-02 23:49:42 UTC
Description of problem:
Sosreport currently does not collect geo-replication master or slave logs located in /var/log/glusterfs/geo-replication and /var/log/glusterfs/geo-replication-slaves

Version-Release number of selected component (if applicable):
sos-2.2-68

How reproducible:
Consistently on gluster nodes with geo-replication configured

Steps to Reproduce:
1. Configure a glusterfs volume as a geo-replication master or slave
2. run sosreport with no flags
3. Resulting tarball does not include geo-replication log files

Actual results:
No geo-rep log files are collected

Expected results:
Recent geo-rep log files are collected

Additional info:
This may be a regression introduced by BZ 1002619

Comment 1 Bryn M. Reeves 2015-02-02 23:58:14 UTC
It grabs the whole of '/var/log/glusterfs' unless log size limits are in effect:

128         if limit:
129             # collect logs last as some of the other actions create log entries
130             self.addCopySpecLimit("/var/log/glusterfs/cli.log", limit)
131             self.addCopySpecLimit("/var/log/glusterfs/*.vol.log", limit)
132             self.addCopySpecLimit("/var/log/glusterfs/gluster-lock.log", limit)
133             self.addCopySpecLimit("/var/log/glusterfs/glustershd.log", limit)
134             self.addCopySpecLimit("/var/log/glusterfs/nfs.log", limit)
135             self.addCopySpecLimit("/var/log/glusterfs/quota-crawl.log", limit)
136             self.addCopySpecLimit("/var/log/glusterfs/quotad.log", limit)
137             self.addCopySpecLimit("/var/log/glusterfs/quotad-mount-*.log", limit)
138             self.addCopySpecLimit("/var/log/glusterfs/status.log", limit)
139             self.addCopySpecLimit("/var/log/glusterfs/bricks/*.log", limit)
140         else:
141             self.addCopySpec("/var/log/glusterfs")

So either pass 'all_logs' or set 'logsize' to zero.

We can easily add these as separate files but that should be reviewed by the Gluster team.

Comment 4 Dustin Black 2015-02-03 00:53:04 UTC
(In reply to Bryn M. Reeves from comment #1)
> It grabs the whole of '/var/log/glusterfs' unless log size limits are in
> effect:
> 
> 128         if limit:
> 129             # collect logs last as some of the other actions create log
> entries
> 130             self.addCopySpecLimit("/var/log/glusterfs/cli.log", limit)
> 131             self.addCopySpecLimit("/var/log/glusterfs/*.vol.log", limit)
> 132             self.addCopySpecLimit("/var/log/glusterfs/gluster-lock.log",
> limit)
> 133             self.addCopySpecLimit("/var/log/glusterfs/glustershd.log",
> limit)
> 134             self.addCopySpecLimit("/var/log/glusterfs/nfs.log", limit)
> 135             self.addCopySpecLimit("/var/log/glusterfs/quota-crawl.log",
> limit)
> 136             self.addCopySpecLimit("/var/log/glusterfs/quotad.log", limit)
> 137            
> self.addCopySpecLimit("/var/log/glusterfs/quotad-mount-*.log", limit)
> 138             self.addCopySpecLimit("/var/log/glusterfs/status.log", limit)
> 139             self.addCopySpecLimit("/var/log/glusterfs/bricks/*.log",
> limit)
> 140         else:
> 141             self.addCopySpec("/var/log/glusterfs")
> 
> So either pass 'all_logs' or set 'logsize' to zero.
> 
> We can easily add these as separate files but that should be reviewed by the
> Gluster team.

Setting 'all_logs' or 'logsize=0' puts us back to the same problem fixed in BZ 1002619 where the collected sosreport can be excessively large. So yes, it's a workaround, but not a desirable one. Better for now it to simply acknowledge that the files are not collected and request them separately. I think it's still bug behavior to not collect these files in a default sosreport run.

I'll move this over to RHS.

Comment 6 Bryn M. Reeves 2015-02-03 07:55:41 UTC
The patch was in the bug for 4 months before release along with a description of what it did (which is accurate) so I can't really view this as a regression (or using limit/all_logs as a workaround).

This is a new feature request: to collect the replication logs explicitly (if this had been mentioned 8 months ago when the changes were proposed they could've been included with bug 1002619).

Comment 7 Vivek Agarwal 2015-12-03 17:21:03 UTC
Thank you for submitting this issue for consideration in Red Hat Gluster Storage. The release for which you requested us to review, is now End of Life. Please See https://access.redhat.com/support/policy/updates/rhs/

If you can reproduce this bug against a currently maintained version of Red Hat Gluster Storage, please feel free to file a new report against the current release.


Note You need to log in before you can comment on or make changes to this bug.