Skip to content
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.

Sign up for GitHub

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

Jump to bottom

rbd-mirror: image live-replay might become stuck #9282

Merged
merged 5 commits into from
May 25, 2016
Merged

rbd-mirror: image live-replay might become stuck #9282

merged 5 commits into from
May 25, 2016

Conversation

dillaman
Copy link

No description provided.

@trociny
Copy link
Contributor

trociny commented May 24, 2016

@dillaman observing this:

zhuzha:~/ceph/ceph.upstream/src% RBD_FEATURES=109 ./ceph_test_librbd --gtest_filter=TestJournalReplay.\*
Note: Google Test filter = TestJournalReplay.*
[==========] Running 13 tests from 1 test case.
[----------] Global test environment set-up.
[----------] 13 tests from TestJournalReplay
[ RUN      ] TestJournalReplay.AioDiscardEvent
test/librbd/journal/test_Replay.cc:169: Failure
Value of: current_tag
  Actual: 1
Expected: initial_tag + 2
Which is: 2
[  FAILED  ] TestJournalReplay.AioDiscardEvent (4615 ms)
[ RUN      ] TestJournalReplay.AioWriteEvent
test/librbd/journal/test_Replay.cc:228: Failure
Value of: current_tag
  Actual: 1
Expected: initial_tag + 2
Which is: 2
[  FAILED  ] TestJournalReplay.AioWriteEvent (522 ms)
[ RUN      ] TestJournalReplay.AioFlushEvent
test/librbd/journal/test_Replay.cc:300: Failure
Value of: current_tag
  Actual: 1
Expected: initial_tag + 2
Which is: 2
[  FAILED  ] TestJournalReplay.AioFlushEvent (755 ms)
[ RUN      ] TestJournalReplay.SnapCreate
[       OK ] TestJournalReplay.SnapCreate (2758 ms)
[ RUN      ] TestJournalReplay.SnapProtect
[       OK ] TestJournalReplay.SnapProtect (2152 ms)
[ RUN      ] TestJournalReplay.SnapUnprotect
[       OK ] TestJournalReplay.SnapUnprotect (2271 ms)
[ RUN      ] TestJournalReplay.SnapRename
[       OK ] TestJournalReplay.SnapRename (1183 ms)
[ RUN      ] TestJournalReplay.SnapRollback
[       OK ] TestJournalReplay.SnapRollback (1053 ms)
[ RUN      ] TestJournalReplay.SnapRemove
[       OK ] TestJournalReplay.SnapRemove (4001 ms)
[ RUN      ] TestJournalReplay.Rename
[       OK ] TestJournalReplay.Rename (619 ms)
[ RUN      ] TestJournalReplay.Resize
[       OK ] TestJournalReplay.Resize (553 ms)
[ RUN      ] TestJournalReplay.Flatten
[       OK ] TestJournalReplay.Flatten (1682 ms)
[ RUN      ] TestJournalReplay.ObjectPosition
[       OK ] TestJournalReplay.ObjectPosition (445 ms)
[----------] 13 tests from TestJournalReplay (22610 ms total)

[----------] Global test environment tear-down
[==========] 13 tests from 1 test case ran. (23759 ms total)
[  PASSED  ] 10 tests.
[  FAILED  ] 3 tests, listed below:
[  FAILED  ] TestJournalReplay.AioDiscardEvent
[  FAILED  ] TestJournalReplay.AioWriteEvent
[  FAILED  ] TestJournalReplay.AioFlushEvent

 3 FAILED TESTS

@dillaman
Copy link
Author

@trociny thanks -- should be better now.

@dillaman
Copy link
Author

@trociny actually, I just moved the unittest failure. fixing.

Jason Dillaman added 5 commits May 24, 2016 12:58
journal: cleanup watch refetch flag handling
ff2cc27 
Clear the refetch required flag while scheduling the watch
and remove the stale object after the watch completes if still
empty. Previously, it was possible for the flag to become
out-of-sync with whether or not it was actually refreshed
and pruned.

Fixes: http://tracker.ceph.com/issues/15993
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
journal: update commit entry object number upon overflow
5fbf5f8 
Otherwise the recorded object positions might point to an older
object that doesn't contain the actual entry.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
journal: keep active tag to assist with pruning watched objects
0fb730e 
It's possible that there might be additional entries to prune in
objects that haven't been prefetched yet. Keep the active tag
to allow these entries to be pruned after they have been loaded.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
journal: eliminate watch delay for object refetches
d7b511b 
The randomized write sizes of the modified rbd-mirror stress
test results in a lot of journal object with few entries.
Immediately fetch objects when performing a refetch check prior
to closing an empty object.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
test: use randomized write sizes for rbd-mirror stress test
b9edff2 
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
@trociny trociny merged commit 20de6de into ceph:master May 25, 2016
@dillaman dillaman deleted the wip-15993 branch May 25, 2016 12:37
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
bug-fix rbd wip-mgolub-testing
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@dillaman @trociny