New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

luminous: osd/PGBackend: delete reply if fails to complete delete request #17233

Merged
merged 2 commits into from Aug 26, 2017

Conversation

Projects
None yet
3 participants
@tchaikov
Contributor

tchaikov commented Aug 24, 2017

if any of the objects fails to be deleted due to pg reset after latest
osdmap, the pg recovery delete reply won't be sent to the primary OSD.
in that case, we should delete the reply.

Fixes: http://tracker.ceph.com/issues/20913
Signed-off-by: Kefu Chai kchai@redhat.com
(cherry picked from commit 1c18b5c)

@tchaikov tchaikov added this to the luminous milestone Aug 24, 2017

@tchaikov

This comment has been minimized.

@liewegas

This comment has been minimized.

Member

liewegas commented Aug 25, 2017

I think this is wrong:

 ceph version 12.1.4-136-g8ac866b (8ac866b55dab9db74abbc68b7f223cd6ad90c45e) luminous (rc)
 1: (()+0xa01bb9) [0x7f05cb0ddbb9]
 2: (()+0x10330) [0x7f05c8be2330]
 3: (gsignal()+0x37) [0x7f05c7c03c37]
 4: (abort()+0x148) [0x7f05c7c07028]
 5: (ceph::__ceph_assert_fail(char const*, char const*, int, char const*)+0x280) [0x7f05cb11a650]
 6: (()+0x40635c) [0x7f05caae235c]
 7: (()+0x52e5da) [0x7f05cac0a5da]
 8: (MOSDPGRecoveryDeleteReply::~MOSDPGRecoveryDeleteReply()+0x38) [0x7f05cadb61a8]
 9: (FunctionContext::finish(int)+0x1a) [0x7f05cac0c4ca]
 10: (Context::complete(int)+0x9) [0x7f05cabf7709]
 11: (C_GatherBase::sub_finish(Context*, int)+0xf7) [0x7f05cadaa777]
 12: (C_GatherBase::C_GatherSub::finish(int)+0x12) [0x7f05cadaa962]
 13: (C_GatherBase::C_GatherSub::complete(int)+0x9) [0x7f05cad70e19]
 14: (()+0x61f6f4) [0x7f05cacfb6f4]
 15: (FunctionContext::finish(int)+0x1a) [0x7f05cac0c4ca]
 16: (Context::complete(int)+0x9) [0x7f05cabf7709]
 17: (std::_Sp_counted_base<(__gnu_cxx::_Lock_policy)2>::_M_release()+0x39) [0x7f05cab8c989]
 18: (ContainerContext >::~ContainerContext()+0x20) [0x7f05caca0e80]
 19: (Finisher::finisher_thread_entry()+0x188) [0x7f05cb117e88]
 20: (()+0x8184) [0x7f05c8bda184]
 21: (clone()+0x6d) [0x7f05c7ccabed]
 NOTE: a copy of the executable, or `objdump -rdS ` is needed to interpret this.

1c18b5c

@tchaikov tchaikov changed the title from luminous: osd/PGBackend: delete reply if fails to complete delete request to [DNM] luminous: osd/PGBackend: delete reply if fails to complete delete request Aug 25, 2017

@tchaikov

This comment has been minimized.

Contributor

tchaikov commented Aug 25, 2017

should also backport #17246 in this PR.

@tchaikov tchaikov changed the title from [DNM] luminous: osd/PGBackend: delete reply if fails to complete delete request to luminous: osd/PGBackend: delete reply if fails to complete delete request Aug 25, 2017

tchaikov added some commits Aug 23, 2017

osd/PGBackend: delete reply if fails to complete delete request
if any of the objects fails to be deleted due to pg reset after latest
osdmap, the pg recovery delete reply won't be sent to the primary OSD.
in that case, we should delete the reply.

Fixes: http://tracker.ceph.com/issues/20913
Signed-off-by: Kefu Chai <kchai@redhat.com>
(cherry picked from commit 1c18b5c)
osd/PGBackend: release a msg using msg->put() not delete
fix the regression introduced by 1c18b5c

Fixes: http://tracker.ceph.com/issues/20913
Signed-off-by: Kefu Chai <kchai@redhat.com>
(cherry picked from commit d2d941d)

@liewegas liewegas merged commit c2ff9fb into ceph:luminous Aug 26, 2017

4 checks passed

Docs: build check OK - docs built
Details
Signed-off-by all commits in this PR are signed
Details
Unmodified Submodules submodules for project are unmodified
Details
make check make check succeeded
Details

@tchaikov tchaikov deleted the tchaikov:wip-luminous-pr-17183 branch Aug 27, 2017

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment