Bug 2356533

Summary: [GSS][Logging] RGW and Ceph-exporter do not resume writing to the log file after being rotated by logrotate.
Product: [Red Hat Storage] Red Hat Ceph Storage Reporter: José Enrique <josgutie>
Component: RGWAssignee: Matt Benjamin (redhat) <mbenjamin>
Status: CLOSED ERRATA QA Contact: Madhavi Kasturi <mkasturi>
Severity: medium Docs Contact:
Priority: medium    
Version: 7.1CC: adking, akane, assingh, bkunal, cbodley, ceph-eng-bugs, cephqe-warriors, dwalveka, gjose, jcaratza, kjosy, ksachdev, mbenjamin, racpatel, rsachere, ver
Target Milestone: ---   
Target Release: 7.1z5   
Hardware: All   
OS: All   
Whiteboard:
Fixed In Version: ceph-18.2.1-340 Doc Type: Bug Fix
Doc Text:
Thread names for bucket logging threads are now set correctly Previously, unsafe thread naming logic added with bucket logging changes allowed well-known threads to have incorrect names. As a result, programs relying on the name of specific threads saw invalid names, and in the special case where an external program (logrotate) was relying on a thread name as a selector, log rotation was disrupted. With this fix, thread names for bucket logging threads are set correctly.
Story Points: ---
Clone Of: Environment:
Last Closed: 2025-06-23 02:51:55 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description José Enrique 2025-04-01 10:25:29 UTC
Description of problem:
Rados Gateway and ceph-exporter loggin doesn't resume after log file are rotated through logrotate.

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

Red Hat Ceph 7.1z3

How reproducible:
execute logrotate --force /etc/logrotate.d/ceph-{fs_id}


Actual results:

The log file is rotated, but the processes do not write to the new log file.

Expected results:

Resume log activity for the mentioned daemons.

Additional info:

The logging activity comes back if the daemons are restarted, but it's not a solution because it's disturb the usual daemon behaviour.

Comment 1 José Enrique 2025-04-04 12:05:32 UTC
New findings,

Using the podman kill function to send the SIGHUP signal, the logging activity comes back, here is the logrotate config used,

# cat /etc/logrotate.d/ceph-0a63c783-54a3-4759-85e9-4b459d3036b4
/var/log/ceph/0a63c783-54a3-4759-85e9-4b459d3036b4/*.log {
    rotate 7
    daily
    compress
    sharedscripts
    postrotate
        podman kill --signal SIGHUP $(podman ps --filter "label=ceph=True" --format "{{.Names}}" | grep -E "^ceph-0a63c783-54a3-4759-85e9-4b459d3036b4") || true
    endscript
    missingok
    notifempty
    su root root
}

Comment 38 errata-xmlrpc 2025-06-23 02:51:55 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory (Red Hat Ceph Storage 7.1 security and bug fix updates), and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHBA-2025:9335

Comment 39 Red Hat Bugzilla 2025-12-20 04:25:03 UTC
The needinfo request[s] on this closed bug have been removed as they have been unresolved for 120 days