This is an automated email from the ASF dual-hosted git repository. granthenke pushed a commit to branch master in repository https://gitbox.apache.org/repos/asf/kudu.git
commit 148c169d46d2cdccc7434e827162552279bc3e4a Author: Andrew Wong <[email protected]> AuthorDate: Wed Mar 4 17:41:32 2020 -0800 test: re-deflake TabletServerQuiescingITest Commit bc2efa1ae97fcc4e592c93d38592206d48d6f8f5 undid some of the deflaking done in 6e4dd49 by increasing the number of read threads. This puts us back at using a single reader, which, while may not always yield active scanners, should be significantly less flaky. I ran this 100 times in TSAN mode and saw no failures, compared to the 20% flakiness reported by our test dashboard. Change-Id: I598c0f0a9a665691bc675920d8dc5a0d272be72e Reviewed-on: http://gerrit.cloudera.org:8080/15365 Reviewed-by: Adar Dembo <[email protected]> Tested-by: Kudu Jenkins --- src/kudu/integration-tests/tablet_server_quiescing-itest.cc | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/src/kudu/integration-tests/tablet_server_quiescing-itest.cc b/src/kudu/integration-tests/tablet_server_quiescing-itest.cc index 921915c..f01cfb1 100644 --- a/src/kudu/integration-tests/tablet_server_quiescing-itest.cc +++ b/src/kudu/integration-tests/tablet_server_quiescing-itest.cc @@ -378,8 +378,9 @@ TEST_F(TServerQuiescingITest, TestQuiescingToolBasics) { auto* ts = cluster_->mini_tablet_server(0); auto rw_workload = CreateFaultIntolerantRWWorkload(); rw_workload->Setup(); - // Spawn a bunch of read threads so we'll be more likely to see scanners. - rw_workload->set_num_read_threads(10); + // NOTE: if this value is too high, this test can become flaky, since the + // degrees of freedom in the number of active scanners will be high. + rw_workload->set_num_read_threads(1); ASSERT_FALSE(ts->server()->quiescing()); const auto& master_addr = cluster_->mini_master()->bound_rpc_addr().ToString(); // First, call the start tool a couple of times.
