test pass as following

sudo ./bin/ceph-objectstore-tool --data-path=./dev/osd0 --pgid 8.7 --op info
{
    "pgid": "8.7",
    "last_update": "29'47",
    "last_complete": "29'47",
    "log_tail": "0'0",
    "last_user_version": 47,
    "last_backfill": "MAX",
    "last_backfill_bitwise": 0,
    "purged_snaps": [],


➜  build git:(wip-19382) ✗ sudo ./bin/ceph-objectstore-tool --data-path=./dev/osd0 --pgid 8.7 --op mark-complete
Marking complete
Marking complete succeeded
➜  build git:(wip-19382) ✗ sudo ./bin/ceph-objectstore-tool --data-path=./dev/osd0 --pgid 8.7 --op info
{
    "pgid": "8.7",
    "last_update": "29'48",
    "last_complete": "29'47",
    "log_tail": "0'0",


➜  build git:(wip-19382) ✗ sudo ./bin/ceph-objectstore-tool --data-path=./dev/osd0 --pgid 8.7 --op reset-last-complete
Rest last_complete
Reset last_complete succeeded
➜  build git:(wip-19382) ✗ sudo ./bin/ceph-objectstore-tool --data-path=./dev/osd0 --pgid 8.7 --op info
{
    "pgid": "8.7",
    "last_update": "29'48",
    "last_complete": "29'48",

@liewegas yes, i prefer to use --force here, because there is an option --force in ceph_objectstore_tool already.

update as following:

➜  build git:(wip-19382) ✗ ./bin/ceph-objectstore-tool --data-path=./dev/osd0 --pgid 4.7 --op reset-last-complete 
WARNING: reset-last-complete is dangerous, to confirm, pass --force option!

jenkins test this please

@liewegas of course! thanks. now this command shows warning message as following:

./bin/ceph-objectstore-tool --data-path=./dev/osd0 --pgid 4.7 --op reset-last-complete
WARNING: reset-last-complete is extremely dangerous and almost certain to lead to permanent data loss unless you know exactly what you are doing. Pass --force to proceed anyway.

"Pass to --force to proceed anyway." changes to "Pass --force to proceed anyway.". is it ok?

@jdurgin Thoughts on this? It makes me really nervous because it seems so much more likely that this will be misused than that it will be used correctly (or ever need to be used correctly). That said, that case is of course guaranteed to present itself shortly after if we don't merge this.

I think this sort of thing is just what ceph-objectstore-tool is for - and it would have helped work around a bug sam encountered (http://tracker.ceph.com/issues/19377). It's probably worth adding something to any ceph-objectstore-tool command that modifies the store to put an audit log of its operation in the objectstore as well.

This pull request has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs.
If you are a maintainer or core committer, please follow-up on this issue to identify what steps should be taken by the author to move this proposed change forward.
If you are the author of this pull request, thank you for your proposed contribution. If you believe this change is still appropriate, please ensure that any feedback has been addressed and ask for a code review.

I think we should merge this. Can you rebase please?

Yeah I will rebase it tomorrow

jenkins test it please

retest this please