Project

General

Profile

Bug #2013

osd: messages for pgs we don't store are never freed

Added by Josh Durgin over 9 years ago. Updated over 9 years ago.

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

0%

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

Description

Once request timestamps are implemented, we could have a timeout period after which misdirected requests are dropped.

Associated revisions

Revision 72bbaeac (diff)
Added by Sage Weil over 9 years ago

osd: discard waiting ops when pg mapping changes

If the pg mapping changes away from us, we can safely discard messages we
have waiting for the PG to be created.

Fixes: #2013
Signed-off-by: Sage Weil <>
Reviewed-by: Josh Durgin <>

History

#1 Updated by Sage Weil over 9 years ago

I think the thing to do is check the waiting map on activate and discard pgs and their messages if they ok longer map to the current node. I don't think we need a timeout?

Well, if we add a tkmeout, it should be part of a general features that tells the client to go away for the current pig until further notice.

#2 Updated by Sage Weil over 9 years ago

  • Status changed from New to In Progress
  • Target version set to v0.43

see wip-pg-waiters?

#3 Updated by Sage Weil over 9 years ago

  • Status changed from In Progress to Resolved

#4 Updated by Sage Weil over 9 years ago

  • Target version changed from v0.43 to v0.42

Also available in: Atom PDF