Hello Andrew Wong, Adar Dembo,
I'd like you to do a code review. Please visit
http://gerrit.cloudera.org:8080/14469
to review the following change.
Change subject: kudu-tool-test: deflake TestFsAddRemoveDataDirEndToEnd
......................................................................
kudu-tool-test: deflake TestFsAddRemoveDataDirEndToEnd
This test was made somewhat flaky by KUDU-2901. The reason is subtle: if any
unrelated IO (such as from a concurrently running test) interleaves with the
creation of the second table, it's possible for the new data directory to be
measured with less available space than the existing ones. The existing LBM
containers have all been preallocated by this point so the test never
remeasures available space. Thus, when the second table flushes, the
KUDU-2901 heuristic ensures that all new data blocks land on the existing
data directories rather than the new one.
The fix is simple: a feature flag for this feature (which we should have had
anyway), which we disable for the test.
Change-Id: Ifb92759318c12d791d26eb8fb6b77914c284f4cb
Reviewed-on: http://gerrit.cloudera.org:8080/14462
Reviewed-by: Andrew Wong <[email protected]>
Tested-by: Adar Dembo <[email protected]>
(cherry picked from commit 7e8f37f8ddf46b965b0af4eba3f31c02b11b8fba)
---
M src/kudu/fs/data_dirs.cc
M src/kudu/tools/kudu-tool-test.cc
2 files changed, 19 insertions(+), 4 deletions(-)
git pull ssh://gerrit.cloudera.org:29418/kudu refs/changes/69/14469/1
--
To view, visit http://gerrit.cloudera.org:8080/14469
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings
Gerrit-Project: kudu
Gerrit-Branch: branch-1.11.x
Gerrit-MessageType: newchange
Gerrit-Change-Id: Ifb92759318c12d791d26eb8fb6b77914c284f4cb
Gerrit-Change-Number: 14469
Gerrit-PatchSet: 1
Gerrit-Owner: Alexey Serbin <[email protected]>
Gerrit-Reviewer: Adar Dembo <[email protected]>
Gerrit-Reviewer: Andrew Wong <[email protected]>
