Project

General

Profile

Actions

Feature #425

closed

trigger osd scrub automatically

Added by Sage Weil over 13 years ago. Updated over 13 years ago.

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

100%

Source:
Tags:
Backport:
Reviewed:
Affected Versions:
Pull request ID:

Description

Not sure if this should be something the mon does (simpler for managing a systemwide scrub policy) or the OSD should do automatically (more scalable). Probably the latter. Any errors will get reported back to the monitor, of course.

Need to decide what constitutdes "the schedule". Just a frequency I guess? g_conf.osd_scrub_interval? Should it (eventually) try to avoid times of high load?


Subtasks 2 (0 open2 closed)

Subtask #485: osd: cooperative scrub schedulingResolvedJosh Durgin10/13/2010

Actions
Subtask #486: osd: make scrub not block writesResolvedSamuel Just10/13/2010

Actions
Actions #1

Updated by Wido den Hollander over 13 years ago

To start with a regular or pseudo-regular interval should be fine, but in larger setups I think a calculation an basis of the load should be more efficient.

And if an OSD triggers the scrub itself, couldn't you get the situation where ALL your OSD's are scrubbing itself thus causing a lot of load inside your cluster?

If could also be a continues process where the mon scrub's all the OSD's one by one, waits for a specified interval and then scrub's them all again.

Actions #2

Updated by Sage Weil over 13 years ago

  • Priority changed from Normal to High
Actions #3

Updated by Sage Weil over 13 years ago

  • Target version changed from v0.22 to v0.23
Actions #4

Updated by Sage Weil over 13 years ago

  • Target version changed from v0.23 to v0.24
Actions #5

Updated by Sage Weil over 13 years ago

  • Assignee set to Josh Durgin
Actions #6

Updated by Sage Weil over 13 years ago

  • Status changed from New to Resolved
Actions

Also available in: Atom PDF