You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ambari.apache.org by "Jayush Luniya (JIRA)" <ji...@apache.org> on 2015/09/26 20:32:04 UTC

[jira] [Updated] (AMBARI-13248) Parallel library should process all futures even if one of them throws an exception

     [ https://issues.apache.org/jira/browse/AMBARI-13248?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jayush Luniya updated AMBARI-13248:
-----------------------------------
    Description: 
- Parallel library should process all futures even if one of them throws an exception
- Consider the case where we spawn 1000 future requests and the very 1st thread throws an exception. 
- Since the exception handling is outside the for loop we would end up returning a result with all nulls to the caller. 
- In getStagesInProgress() as a fall back we check the ParallelLoopResult and get all the missing results sequentially, so this could lead to the main thread and the spawned threads will process same requests in parallel and could lead to issues.

> Parallel library should process all futures even if one of them throws an exception
> -----------------------------------------------------------------------------------
>
>                 Key: AMBARI-13248
>                 URL: https://issues.apache.org/jira/browse/AMBARI-13248
>             Project: Ambari
>          Issue Type: Bug
>          Components: ambari-agent, ambari-server
>    Affects Versions: 2.1.2
>            Reporter: Jayush Luniya
>            Assignee: Jayush Luniya
>            Priority: Blocker
>             Fix For: 2.1.2
>
>
> - Parallel library should process all futures even if one of them throws an exception
> - Consider the case where we spawn 1000 future requests and the very 1st thread throws an exception. 
> - Since the exception handling is outside the for loop we would end up returning a result with all nulls to the caller. 
> - In getStagesInProgress() as a fall back we check the ParallelLoopResult and get all the missing results sequentially, so this could lead to the main thread and the spawned threads will process same requests in parallel and could lead to issues.



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