Project

General

Profile

Actions

Bug #46140

closed

mds: couldn't see the logs in log file before the daemon get aborted

Added by Xiubo Li almost 4 years ago. Updated over 1 year ago.

Status:
Resolved
Priority:
Normal
Category:
-
Target version:
-
% Done:

0%

Source:
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Component(FS):
Labels (FS):
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

It seems the assert() call doesn't flush the log buffer to the relevant log files before it aborting the daemons, usually the logs are very useful to diagnose the root cause of the problem.

Actions #1

Updated by Dhairya Parmar over 1 year ago

  • Assignee changed from Xiubo Li to Dhairya Parmar
Actions #2

Updated by Xiubo Li over 1 year ago

  • Description updated (diff)
Actions #3

Updated by Dhairya Parmar over 1 year ago

  • Status changed from New to Closed

After a brief discussion with @Xiubo Li, we decided to close this tracker as this issue was encountered while debugging a crash bug long time ago and since then it has not been seen/reported. It should be a Log thread that got stuck when flushing log buffer to the log files.

Actions #4

Updated by Xiubo Li over 1 year ago

  • Description updated (diff)
Actions #5

Updated by Xiubo Li over 1 year ago

I recalled it, we need to switch `assert()` to `ceph_assert()`. And the `ceph_assert()` will help dump the recent logs from the log buffer. And also the `assert()` is buggy in ceph, more detail please see ceph PR#42541.

Actions #6

Updated by Xiubo Li over 1 year ago

  • Status changed from Closed to New
Actions #7

Updated by Xiubo Li over 1 year ago

  • Status changed from New to Resolved

Checked the code, all the assert()/abort() have been fixed. Closing it.

Actions

Also available in: Atom PDF