HBASE-27720 TestClusterRestartFailover is flakey due to metrics asser…

…tion (apache#5110) Signed-off-by: Duo Zhang <[email protected]> (cherry picked from commit f36587c) (cherry picked from commit 34e05b7) Change-Id: Ide31e7d4473061e223752c213c6836f5a53a1979
vinayakphegde · Apr 8, 2023 · fa67a90 · fa67a90
1 parent 7644b73
commit fa67a90
Showing 1 changed file with 6 additions and 0 deletions.
diff --git a/hbase-server/src/test/java/org/apache/hadoop/hbase/master/TestClusterRestartFailover.java b/hbase-server/src/test/java/org/apache/hadoop/hbase/master/TestClusterRestartFailover.java
@@ -151,6 +151,12 @@ private void setupCluster() throws Exception {
     LOG.info("Setup cluster");
     UTIL.startMiniCluster(StartMiniClusterOption.builder().masterClass(HMasterForTest.class)
       .numMasters(1).numRegionServers(3).build());
+    // this test has been flaky. When it is rerun by surefire, the underlying minicluster isn't
+    // completely cleaned. specifically, the metrics system isn't reset. The result is an otherwise
+    // successful re-run is failed because there's 8 or 12 SCPcounts instead of the 4 that a
+    // single run of the test would otherwise produce. Thus, explicitly reset the metrics source
+    // each time we setup the cluster.
+    UTIL.getMiniHBaseCluster().getMaster().getMasterMetrics().getMetricsSource().init();
     LOG.info("Cluster is up");
     UTIL.waitFor(60000, () -> UTIL.getMiniHBaseCluster().getMaster().isInitialized());
     LOG.info("Master is up");