Siyao Meng created HDDS-17077:
---------------------------------
Summary: Intermittent failure in
TestKeyDeletingService$Normal#checkIfDeleteServiceIsDeletingKeys
Key: HDDS-17077
URL: https://issues.apache.org/jira/browse/HDDS-17077
Project: Apache Ozone
Issue Type: Sub-task
Components: Ozone Manager, test
Reporter: Siyao Meng
Assignee: Siyao Meng
{code}
Tests run: 12, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 33.23 s <<<
FAILURE! -- in org.apache.hadoop.ozone.om.service.TestKeyDeletingService$Normal
org.apache.hadoop.ozone.om.service.TestKeyDeletingService$Normal.checkIfDeleteServiceIsDeletingKeys
-- Time elapsed: 0.616 s <<< FAILURE!
java.lang.AssertionError:
Expecting empty but was:
{"/checkIfDeleteServiceIsDeletingKeys/bucket0000000772/key0000000773/-9223372036854628608/0"=PurgedKey{...}}
at
org.apache.hadoop.ozone.om.service.TestKeyDeletingService$Normal.checkIfDeleteServiceIsDeletingKeys(TestKeyDeletingService.java:258)
{code}
h3. Root cause
{{KeyDeletingService}} increments {{deletedKeyCount}} as soon as the purge
request is applied, but the deletedTable entries are removed from RocksDB only
when the OM double buffer flushes. The test waits for {{deletedKeyCount}} to
reach 100 and then iterates the deletedTable through
{{getPendingDeletionKeys}}, which reads the DB. In the failing run the last
service run purged a single key ({{key0000000773}}) and the test read the table
before that purge was flushed.
{{checkDeletedTableCleanUpForSnapshot}} has the same check after the same kind
of wait.
h3. Fix
Call {{om.awaitDoubleBufferFlush()}} before reading the deletedTable in both
tests.
- https://github.com/apache/ozone/actions/runs/38935632776/job/117106007880
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
