[ https://issues.apache.org/jira/browse/CASSANDRA-14603?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Sam Tunnicliffe updated CASSANDRA-14603: ---------------------------------------- Fix Version/s: 3.0.x 2.2.x Status: Patch Available (was: Open) The problem is with the tests - when I updated these for CASSANDRA-14134 I forgot that value skipping during reads was disabled until CASSANDRA-10657 re-introduced in 3.4 PR for the dtest fixes: https://github.com/apache/cassandra-dtest/pull/33 CI runs with the fix: [2.2|https://circleci.com/gh/beobal/cassandra/250], [3.0|https://circleci.com/gh/beobal/cassandra/252], [3.11|https://circleci.com/gh/beobal/cassandra/252] > [dtest] read_repair_test.TestReadRepair > --------------------------------------- > > Key: CASSANDRA-14603 > URL: https://issues.apache.org/jira/browse/CASSANDRA-14603 > Project: Cassandra > Issue Type: Bug > Components: Testing > Reporter: Jason Brown > Assignee: Sam Tunnicliffe > Priority: Major > Labels: dtest > Fix For: 2.2.x, 3.0.x > > > tests {{test_alter_rf_and_run_read_repair}} and {{test_read_repair_chance}} > consistently fail on 3.0; the latter also fails on 2.2. I suspect it's the > same cause, as the output from pytest shows the same error in the same shared > function ({{check_data_on_each_replica}}): > {noformat} > res = rows_to_list(session.execute(stmt)) > logger.debug("Actual result: " + str(res)) > expected = [[1, 1, 1]] if expect_fully_repaired or n == > initial_replica else [[1, 1, None]] > if res != expected: > > raise NotRepairedException() > E read_repair_test.NotRepairedException > read_repair_test.py:204: NotRepairedException > {noformat} -- This message was sent by Atlassian JIRA (v7.6.3#76005) --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org