You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@storm.apache.org by "Rick Kellogg (JIRA)" <ji...@apache.org> on 2015/10/05 03:05:27 UTC

[jira] [Updated] (STORM-674) Add chaos monkey testing to storm

     [ https://issues.apache.org/jira/browse/STORM-674?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Rick Kellogg updated STORM-674:
-------------------------------
    Component/s: storm-core

> Add chaos monkey testing to storm
> ---------------------------------
>
>                 Key: STORM-674
>                 URL: https://issues.apache.org/jira/browse/STORM-674
>             Project: Apache Storm
>          Issue Type: New JIRA Project
>          Components: storm-core
>            Reporter: Parth Brahmbhatt
>
> Storm is a distributed processing system and  as we add new features like Nimbus HA  its not always easy to test the correctness and robustness of storm under failure scenarios across multiple hosts.Ideally storm will have a suit of chaos monkey tests that can be executed to gain confidence that any new features/ or changes to these new features in future has not affected storm's robustness. 
> We could add simple kill, restart(numSecsToWait) thrift APIs that can be called  remotely to kill or restart nimbus/supervisor/workers to begin with. Alternatively we could evaluate some frameworks out there for ingesting host/network failures.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)