[
https://issues.apache.org/jira/browse/CASSANDRA-17304?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Josh McKenzie updated CASSANDRA-17304:
--------------------------------------
Bug Category: Parent values: Correctness(12982)Level 1 values: Test
Failure(12990)
> Test Failure:
> dtest.snapshot_test.TestArchiveCommitlog.test_dont_archive_commitlog
> ----------------------------------------------------------------------------------
>
> Key: CASSANDRA-17304
> URL: https://issues.apache.org/jira/browse/CASSANDRA-17304
> Project: Cassandra
> Issue Type: Bug
> Components: Test/dtest/python
> Reporter: Josh McKenzie
> Priority: Normal
>
> https://ci-cassandra.apache.org/job/Cassandra-4.0/321/testReport/dtest.snapshot_test/TestArchiveCommitlog/test_dont_archive_commitlog/
> Failed 1 times in the last 30 runs. Flakiness: 3%, Stability: 96%
> Error Message
> AssertionError: It's been over a {s}s and we haven't written a new commitlog
> segment. Something is wrong.
> {code}
> Stacktrace
> self = <snapshot_test.TestArchiveCommitlog object at 0x7feb0a001e80>
> def test_dont_archive_commitlog(self):
> """
> Run the archive commitlog test, but forget to add the restore
> commands
> """
> > self.run_archive_commitlog(restore_point_in_time=False,
> > restore_archived_commitlog=False)
> snapshot_test.py:253:
> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
> _
> snapshot_test.py:300: in run_archive_commitlog
> advance_to_next_cl_segment(
> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
> _
> session = <cassandra.cluster.Session object at 0x7feb09f861c0>
> commitlog_dir =
> '/home/cassandra/cassandra/cassandra-dtest/tmp/dtest-1mc01m0r/test/node1/commitlogs'
> keyspace_name = 'ks', table_name = 'junk_table', timeout = 120
> def advance_to_next_cl_segment(session, commitlog_dir,
> keyspace_name='ks',
> table_name='junk_table',
> timeout=120):
> """
> This is a hack to work around problems like CASSANDRA-11811.
>
> The problem happens in commitlog-replaying tests, like the snapshot
> and CDC
> tests. If we replay the first commitlog that's created, we wind up
> replaying some mutations that initialize system tables, so this
> function
> advances the node to the next CL by filling up the first one.
> """
> session.execute(
> 'CREATE TABLE {ks}.{tab} ('
> 'a uuid PRIMARY KEY, b uuid, c uuid, d uuid, '
> 'e uuid, f uuid, g uuid, h uuid'
> ')'.format(ks=keyspace_name, tab=table_name)
> )
> prepared_insert = session.prepare(
> 'INSERT INTO {ks}.{tab} '
> '(a, b, c, d, e, f, g, h) '
> 'VALUES ('
> 'uuid(), uuid(), uuid(), uuid(), '
> 'uuid(), uuid(), uuid(), uuid()'
> ')'.format(ks=keyspace_name, tab=table_name)
> )
>
> # record segments that we want to advance past
> initial_cl_files = _files_in(commitlog_dir)
>
> start = time.time()
> stop_time = start + timeout
> rate_limited_debug_logger = get_rate_limited_function(logger.debug, 5)
> logger.debug('attempting to write until we start writing to new CL
> segments: {}'.format(initial_cl_files))
>
> while _files_in(commitlog_dir) <= initial_cl_files:
> elapsed = time.time() - start
> rate_limited_debug_logger(' commitlog-advancing load step has
> lasted {s:.2f}s'.format(s=elapsed))
> > assert (
> time.time() <= stop_time), "It's been over a {s}s and we
> haven't written a new " + \
> "commitlog segment. Something is wrong.".format(s=timeout)
> E AssertionError: It's been over a {s}s and we haven't written a
> new commitlog segment. Something is wrong.
> tools/hacks.py:59: AssertionError
> {code}
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
