Skip to content

octopus: bluestore: Rescue procedure for extremely large bluefs log#36123

Merged
yuriw merged 3 commits intoceph:octopusfrom
aclamk:wip-bluefs-log-replay-rescue-octopus
Aug 10, 2020
Merged

octopus: bluestore: Rescue procedure for extremely large bluefs log#36123
yuriw merged 3 commits intoceph:octopusfrom
aclamk:wip-bluefs-log-replay-rescue-octopus

Conversation

@aclamk
Copy link
Contributor

@aclamk aclamk commented Jul 15, 2020

backport tracker: https://tracker.ceph.com/issues/46599

backport of: #36121

parent tracker: https://tracker.ceph.com/issues/46552


Original PR description:

This is procedure that stems from https://tracker.ceph.com/issues/46195.
It was created and tested in Luminous, but now needs to be ported to all other ceph branches.
This one is a cherry-picked from #36110 #36121
It fixes https://tracker.ceph.com/issues/46552

aclamk added 3 commits July 15, 2020 22:14
This is a procedure tries to find on disk unreachable extents and pretend they were
already a part of bluefs log. If this gives proper crc, accept it.

Fixes: https://tracker.ceph.com/issues/46552
Signed-off-by: Adam Kupczyk <akupczyk@redhat.com>
(cherry-picked from commit 40e94a1)

Conflicts:
	src/os/bluestore/BlueFS.cc
This ability only makes sense as a step that allows to perform fsck before commiting recovered bluefs log.

Fixes: https://tracker.ceph.com/issues/46552
Signed-off-by: Adam Kupczyk <akupczyk@redhat.com>
(cherry-picked from commit 7b01af4)

Conflicts:
	src/os/bluestore/BlueFS.cc
Adds additional paragraph to ceph-bluestore-tool documentation,
describing how to use *special* options --bluefs_replay_recovery
and --bluefs_replay_recovery_disable_compact to recover large
bluefs log.

Fixes: https://tracker.ceph.com/issues/46552
Signed-off-by: Adam Kupczyk <akupczyk@redhat.com>
(cherry-picked from commit d7a49b0)
@aclamk aclamk force-pushed the wip-bluefs-log-replay-rescue-octopus branch from 4ff901d to c861dc1 Compare July 15, 2020 20:17
@smithfarm smithfarm changed the title Octopus: Rescue procedure for extremely large bluefs log octopus: bluestore: Rescue procedure for extremely large bluefs log Jul 17, 2020
@smithfarm smithfarm added this to the octopus milestone Jul 17, 2020
Copy link
Contributor

@smithfarm smithfarm left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Backport PR opened prematurely. The "cherry picked from commit ..." lines in the commit messages contain SHA1s that are not in master. If the master PR containing these commits - #36121 - gets rebased or force-pushed before it is merged, this PR will have to be redone.

@smithfarm smithfarm added the DNM label Jul 17, 2020
@smithfarm
Copy link
Contributor

DNM until master PR merges and SHA1s are checked.

@neha-ojha
Copy link
Member

DNM until master PR merges and SHA1s are checked.

@smithfarm The master PR has now merged and SHA1s look right to me.

@yuriw
Copy link
Contributor

yuriw commented Aug 4, 2020

@yuriw yuriw merged commit dbd7a1b into ceph:octopus Aug 10, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants