[
https://issues.apache.org/jira/browse/TEZ-119?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14348220#comment-14348220
]
Jeff Zhang commented on TEZ-119:
--------------------------------
Looks like it has been done.
TEZ_AM_RM_HEARTBEAT_INTERVAL_MS_MAX is used for this.
See YarnTaskSchedulerService.java
{code}
int heartbeatIntervalMax = conf.getInt(
TezConfiguration.TEZ_AM_RM_HEARTBEAT_INTERVAL_MS_MAX,
TezConfiguration.TEZ_AM_RM_HEARTBEAT_INTERVAL_MS_MAX_DEFAULT);
amRmClient.setHeartbeatInterval(heartbeatIntervalMax);
{code}
> The AM-RM heartbeat interval should not be static
> -------------------------------------------------
>
> Key: TEZ-119
> URL: https://issues.apache.org/jira/browse/TEZ-119
> Project: Apache Tez
> Issue Type: Improvement
> Reporter: Siddharth Seth
> Assignee: Bikas Saha
>
> AMs should be more aggressive in heartbeating to the RM - especially soon
> after job start to get the initial set of containers, and also in the general
> case where allocations are pending.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
