[YARN-6136] YARN registry service should avoid scanning whole ZK tree for every container/application finish

• Type: Sub-task
• Status: Resolved
• Priority: Critical
• Resolution: Invalid
• Component/s: api, resourcemanager
• Assignee: Wangda Tan
• Reporter: Wangda Tan
• Created: 31/Jan/17 21:53
• Updated: 22/Mar/18 18:20
• Resolved: 22/Mar/18 18:20

Description

In existing registry service implementation, purge operation triggered by container finish event:

  public void onContainerFinished(ContainerId id) throws IOException {
    LOG.info("Container {} finished, purging container-level records",
        id);
    purgeRecordsAsync("/",
        id.toString(),
        PersistencePolicies.CONTAINER);
  }

Since this happens on every container finish, so it essentially scans all (or almost) ZK node from the root.

We have a cluster which have hundreds of ZK nodes for service registry, and have 20K+ ZK nodes for other purposes. The existing implementation could generate massive ZK operations and internal Java objects (RegistryPathStatus) as well. The RM becomes very unstable when there're batch container finish events because of full GC pause and ZK connection failure.

Issue Links

• is caused by: YARN-2571 RM to support YARN registry