[
https://issues.apache.org/jira/browse/TEZ-2631?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jeff Zhang updated TEZ-2631:
----------------------------
Description:
The ChaosMonkey is a daemon thread that can kill AM randomly. It could be
enabled through TezConfiguration. The ChaosMonkey thread can kill the AM
evenly in the lifecycle of one dag.
The initial idea is that this daemon thread will query the progress of dag
periodically, and compare the progress with an random number. If the gap is
between a threshold (could also be randomly), then kill AM.
> Add ChaosMonkey in AM for ChaosMonkey for recovery system test
> --------------------------------------------------------------
>
> Key: TEZ-2631
> URL: https://issues.apache.org/jira/browse/TEZ-2631
> Project: Apache Tez
> Issue Type: Bug
> Reporter: Jeff Zhang
>
> The ChaosMonkey is a daemon thread that can kill AM randomly. It could be
> enabled through TezConfiguration. The ChaosMonkey thread can kill the AM
> evenly in the lifecycle of one dag.
> The initial idea is that this daemon thread will query the progress of dag
> periodically, and compare the progress with an random number. If the gap is
> between a threshold (could also be randomly), then kill AM.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
