ANANDA G B created YARN-9674:
Summary: 'Max AM Resource' calculated for default partition using
'Effective Max Capacity' and ohter partitions it using 'Effective Capacity'
Key: YARN-9674
URL:
[
https://issues.apache.org/jira/browse/YARN-9673?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
chan updated YARN-9673:
---
Description:
We have 1000 nodes in the cluster. Recently I found that when many tasks are
submitted to the
[
https://issues.apache.org/jira/browse/YARN-9673?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
chan updated YARN-9673:
---
Environment: (was: protected void handleStoreEvent(RMStateStoreEvent
event) {
this.writeLock.lock();
try {
if
[
https://issues.apache.org/jira/browse/YARN-9673?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
chan updated YARN-9673:
---
Description:
We have 1000 nodes in the cluster. Recently I found that when many tasks are
submitted to the
chan created YARN-9673:
--
Summary: RMStateStore writeLock make app waste more time
Key: YARN-9673
URL: https://issues.apache.org/jira/browse/YARN-9673
Project: Hadoop YARN
Issue Type: Bug
