Viraj Jasani created HDFS-16947:
-----------------------------------
Summary: RBF NamenodeHeartbeatService to report error for not
being able to register namenode in state store
Key: HDFS-16947
URL: https://issues.apache.org/jira/browse/HDFS-16947
Project: Hadoop HDFS
Issue Type: Improvement
Reporter: Viraj Jasani
Assignee: Viraj Jasani
Namenode heartbeat service should provide error with full stacktrace if it
cannot register namenode in the state store. As of today, we only log info msg.
For zookeeper based impl, this might mean either a) curator manager is not
initialized or b) if it fails to write to znode after exhausting retries. For
either of these cases, reporting only INFO log might not be good enough and we
might have to look for errors elsewhere.
Sample example:
{code:java}
2023-02-20 23:10:33,714 DEBUG [NamenodeHeartbeatService {ns} nn0-0]
router.NamenodeHeartbeatService - Received service state: ACTIVE from HA
namenode: {ns}-nn0:nn-0-{ns}.{cluster}:9000
2023-02-20 23:10:33,731 INFO [NamenodeHeartbeatService {ns} nn0-0]
impl.MembershipStoreImpl - Inserting new NN registration:
nn-0.namenode.{cluster}:8888->{ns}:nn0:nn-0-{ns}.{cluster}:9000-ACTIVE
2023-02-20 23:10:33,731 INFO [NamenodeHeartbeatService {ns} nn0-0]
router.NamenodeHeartbeatService - Cannot register namenode in the State Store
{code}
If we could log full stacktrace:
{code:java}
2023-02-21 00:20:24,691 ERROR [NamenodeHeartbeatService {ns} nn0-0]
router.NamenodeHeartbeatService - Cannot register namenode in the State Store
org.apache.hadoop.hdfs.server.federation.store.StateStoreUnavailableException:
State Store driver StateStoreZooKeeperImpl in nn-0.namenode.{cluster} is not
ready.
at
org.apache.hadoop.hdfs.server.federation.store.driver.StateStoreDriver.verifyDriverReady(StateStoreDriver.java:158)
at
org.apache.hadoop.hdfs.server.federation.store.driver.impl.StateStoreZooKeeperImpl.putAll(StateStoreZooKeeperImpl.java:235)
at
org.apache.hadoop.hdfs.server.federation.store.driver.impl.StateStoreBaseImpl.put(StateStoreBaseImpl.java:74)
at
org.apache.hadoop.hdfs.server.federation.store.impl.MembershipStoreImpl.namenodeHeartbeat(MembershipStoreImpl.java:179)
at
org.apache.hadoop.hdfs.server.federation.resolver.MembershipNamenodeResolver.registerNamenode(MembershipNamenodeResolver.java:381)
at
org.apache.hadoop.hdfs.server.federation.router.NamenodeHeartbeatService.updateState(NamenodeHeartbeatService.java:317)
at
org.apache.hadoop.hdfs.server.federation.router.NamenodeHeartbeatService.lambda$periodicInvoke$0(NamenodeHeartbeatService.java:244)
...
... {code}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]