[ 
https://issues.apache.org/jira/browse/YARN-7565?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16297775#comment-16297775
 ] 

Chandni Singh commented on YARN-7565:
-------------------------------------

[~eyang] In the recovery, the service record is fetched from the Registry. 
In ComponentInstance record.description is assigned the compInstanceName.
{code}
 // Write service record into registry
  private  void updateServiceRecord(
      YarnRegistryViewForProviders yarnRegistry, ContainerStatus status) {
    ServiceRecord record = new ServiceRecord();
    String containerId = status.getContainerId().toString();
    record.set(YARN_ID, containerId);
    record.description = getCompInstanceName();
    record.set(YARN_PERSISTENCE, PersistencePolicies.CONTAINER);
    record.set(YARN_IP, status.getIPs().get(0));
    record.set(YARN_HOSTNAME, status.getHost());
    record.set(YARN_COMPONENT, component.getName());
    try {
      yarnRegistry
          .putComponent(RegistryPathUtils.encodeYarnID(containerId), record);
    } catch (IOException e) {
      LOG.error(
          "Failed to update service record in registry: " + containerId + "");
    }
  }
{code}

> Yarn service pre-maturely releases the container after AM restart 
> ------------------------------------------------------------------
>
>                 Key: YARN-7565
>                 URL: https://issues.apache.org/jira/browse/YARN-7565
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>            Reporter: Chandni Singh
>            Assignee: Chandni Singh
>             Fix For: 3.1.0
>
>         Attachments: YARN-7565.001.patch, YARN-7565.002.patch, 
> YARN-7565.003.patch, YARN-7565.004.patch, YARN-7565.005.patch, 
> YARN-7565.addendum.001.patch
>
>
> With YARN-6168, recovered containers can be reported to AM in response to the 
> AM heartbeat. 
> Currently, the Service Master will release the containers, that are not 
> reported in the AM registration response, immediately.
> Instead, the master can wait for a configured amount of time for the 
> containers to be recovered by RM. These containers are sent to AM in the 
> heartbeat response. Once a container is not reported in the configured 
> interval, it can be released by the master.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org

Reply via email to