Fix assertion failure in IndexShard.updateGlobalCheckpointOnReplica()…

… when remote translog is enabled (#6975) Signed-off-by: Sachin Kale <[email protected]>
opensearch-project · Apr 4, 2023 · 95c6ed9 · 95c6ed9
1 parent e12a5b9
commit 95c6ed9
Show file tree

Hide file tree

Showing 2 changed files with 4 additions and 6 deletions.
diff --git a/server/src/internalClusterTest/java/org/opensearch/remotestore/RemoteStoreIT.java b/server/src/internalClusterTest/java/org/opensearch/remotestore/RemoteStoreIT.java
@@ -282,22 +282,18 @@ public void testPeerRecoveryWithRemoteStoreNoRemoteTranslogRefresh() throws Exce
         testPeerRecovery(false, randomIntBetween(2, 5), false);
     }
 
-    @AwaitsFix(bugUrl = "https://github.com/opensearch-project/OpenSearch/issues/6193")
     public void testPeerRecoveryWithRemoteStoreAndRemoteTranslogNoDataFlush() throws Exception {
         testPeerRecovery(true, 1, true);
     }
 
-    @AwaitsFix(bugUrl = "https://github.com/opensearch-project/OpenSearch/issues/6193")
     public void testPeerRecoveryWithRemoteStoreAndRemoteTranslogFlush() throws Exception {
         testPeerRecovery(true, randomIntBetween(2, 5), true);
     }
 
-    @AwaitsFix(bugUrl = "https://github.com/opensearch-project/OpenSearch/issues/6193")
     public void testPeerRecoveryWithRemoteStoreAndRemoteTranslogNoDataRefresh() throws Exception {
         testPeerRecovery(true, 1, false);
     }
 
-    @AwaitsFix(bugUrl = "https://github.com/opensearch-project/OpenSearch/issues/6193")
     public void testPeerRecoveryWithRemoteStoreAndRemoteTranslogRefresh() throws Exception {
         testPeerRecovery(true, randomIntBetween(2, 5), false);
     }

diff --git a/server/src/main/java/org/opensearch/index/shard/IndexShard.java b/server/src/main/java/org/opensearch/index/shard/IndexShard.java
@@ -3078,9 +3078,11 @@ public void updateGlobalCheckpointOnReplica(final long globalCheckpoint, final S
              * calculations of the global checkpoint. However, we can not assert that we are in the translog stage of recovery here as
              * while the global checkpoint update may have emanated from the primary when we were in that state, we could subsequently move
              * to recovery finalization, or even finished recovery before the update arrives here.
+             * When remote translog is enabled for an index, replication operation is limited to primary term validation and does not
+             * update local checkpoint at replica, so the local checkpoint at replica can be less than globalCheckpoint.
              */
-            assert state() != IndexShardState.POST_RECOVERY && state() != IndexShardState.STARTED
-                : "supposedly in-sync shard copy received a global checkpoint ["
+            assert (state() != IndexShardState.POST_RECOVERY && state() != IndexShardState.STARTED)
+                || indexSettings.isRemoteTranslogStoreEnabled() : "supposedly in-sync shard copy received a global checkpoint ["
                     + globalCheckpoint
                     + "] "
                     + "that is higher than its local checkpoint ["