Project

General

Profile

Actions
Copy link

Feature #51716

closed

Add option in `fs new` command to start rank 0 in failed state

Added by Ramana Raja almost 3 years ago. Updated over 1 year ago.

Status:
Resolved
Priority:
Normal
Assignee:
Ramana Raja
Category:
-
Target version:
Ceph - v17.0.0
% Done:

0%

Source:
Development
Tags:
backport_processed
Backport:
pacific
Reviewed:
Affected Versions:
Component(FS):
MDSMonitor
Labels (FS):
Pull request ID:
42763

Description

Source: https://github.com/ceph/ceph/pull/42295#discussion_r670827459

Currently, to recover a file system after recovering monitor store, you need to stop all the MDSs, create FSMap with defaults using `fs new` command, execute `fs reset` command to get file system' rank 0 into failed state, and then restart MDSs.

If the rank 0 of filesystem is not set to failed state, then the MDS that takes up rank 0 doesn't read the existing in RADOS metadata and starts overwriting it. So the fiile system is not recovered. By allowing the `fs new` command to optionally set rank 0 to failed state instead of creating state, we may able to avoid the need to stop all the MDSs and execute `fs reset` command.

Related issues 2 (1 open — 1 closed)

Related to CephFS - Tasks #51341: Steps to recover file system(s) after recovering the Ceph monitor storeIn Progress

Actions
Copied to CephFS - Backport #52680: pacific: Add option in `fs new` command to start rank 0 in failed stateResolvedRamana RajaActions
Actions
Copy link
 #1

Updated by Ramana Raja almost 3 years ago

  • Subject changed from Add option to start rank 0 in failed state for `fs new` command to Add option in `fs new` command to start rank 0 in failed state
Actions
Copy link
 #2

Updated by Ramana Raja almost 3 years ago

  • Related to Tasks #51341: Steps to recover file system(s) after recovering the Ceph monitor store added
Actions
Copy link
 #3

Updated by Patrick Donnelly almost 3 years ago

Another thing I thought of after our discussion today, Ramana: I think the --recover flag should do:

- Set rank0 to failed.
- Set the new fs to not be joinable (so MDS do not need turned off)

Actions
Copy link
 #4

Updated by Ramana Raja almost 3 years ago

Patrick Donnelly wrote:

Another thing I thought of after our discussion today, Ramana: I think the --recover flag should do:

- Set rank0 to failed.
- Set the new fs to not be joinable (so MDS do not need turned off)

Thanks, Patrick. I will try it out.

Actions
Copy link
 #5

Updated by Ramana Raja over 2 years ago

  • Status changed from New to In Progress
  • Pull request ID set to 42763
Actions
Copy link
 #6

Updated by Patrick Donnelly over 2 years ago

  • Status changed from In Progress to Fix Under Review
  • Assignee set to Ramana Raja
  • Target version set to v17.0.0
  • Source set to Development
  • Backport set to pacific
Actions
Copy link
 #7

Updated by Patrick Donnelly over 2 years ago

  • Status changed from Fix Under Review to Pending Backport
Actions
Copy link
 #8

Updated by Backport Bot over 2 years ago

  • Copied to Backport #52680: pacific: Add option in `fs new` command to start rank 0 in failed state added
Actions
Copy link
 #9

Updated by Backport Bot over 1 year ago

  • Tags set to backport_processed
Actions
Copy link
 #10

Updated by Patrick Donnelly over 1 year ago

  • Status changed from Pending Backport to Resolved
Actions
Copy link

Also available in: Atom PDF