You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Stephan Ewen (JIRA)" <ji...@apache.org> on 2015/08/06 19:13:04 UTC
[jira] [Resolved] (FLINK-266) Warn user if cluster did not come up
as expected
[ https://issues.apache.org/jira/browse/FLINK-266?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Stephan Ewen resolved FLINK-266.
--------------------------------
Resolution: Won't Fix
Fix Version/s: (was: pre-apache)
The deployment model is different now, the pre-flight phase and optimizer do not connect to the cluster any more to gather the availability of resources. Therefor, this issue is not really an issue any more ;-)
> Warn user if cluster did not come up as expected
> ------------------------------------------------
>
> Key: FLINK-266
> URL: https://issues.apache.org/jira/browse/FLINK-266
> Project: Flink
> Issue Type: Improvement
> Reporter: GitHub Import
> Labels: github-import
>
> While I did some work on a cluster, I was wondering why my job did not utilize all TaskManagers.
> It seems that I started my job too early (before all TaskManager registered with the JobManager) and therefore, the compiler did not consider them.
> We should either make the `start-cluster.sh` script blocking (with a timeout). Or the pact-client.sh should report a warning if less TaskManagers than expected (number in `slaves` file) are up.
> ---------------- Imported from GitHub ----------------
> Url: https://github.com/stratosphere/stratosphere/issues/266
> Created by: [rmetzger|https://github.com/rmetzger]
> Labels: enhancement, runtime, simple-issue,
> Created at: Mon Nov 11 15:56:56 CET 2013
> State: open
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)