Project

General

Profile

Actions
Copy link

Bug #8341

open

improve falover to next available MON

Added by Dmitry Smirnov almost 10 years ago.

Status:
New
Priority:
Normal
Assignee:
-
Category:
-
Target version:
-
% Done:

0%

Source:
Community (user)
Tags:
Backport:
Regression:
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Crash signature (v1):
Crash signature (v2):

Description

I suspend one computer with RBD mapped device fairly often.
Every time it comes out of suspend gracefully but today it didn't -- RBD device stayed unresponsive after resume.
Investigation revealed one of four monitors was down at the time of wake up.
Starting affected monitor immediately revived RBD device.

Normally taking any monitor down do not affect RBD devices responsiveness as failover happen quite fast.
I'm not sure why RBD client did not switch to another monitor after wake-up.
I hope failover to next available MON can be improved.

This incident happened with just released ceph_0.80.1 and latest ceph-client from "for-linus" branch + patch from #8226 (built as DKMS module, running on Linux-3.14.2.

No data to display

Actions
Copy link

Also available in: Atom PDF