flink-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From " Mario Georgiev (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (FLINK-11900) Flink on Kubernetes sensitive about arguments placement
Date Wed, 13 Mar 2019 11:07:00 GMT

     [ https://issues.apache.org/jira/browse/FLINK-11900?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Mario Georgiev updated FLINK-11900:
-----------------------------------
    Description: 
Hello guys,

I've discovered that when deploying the job cluster on Kubernetes, the Job Cluster Manager
seems sensitive about the placement of arguments. 

For instance if i put the savepoint argument last, it never reads it. 

For instance if arguments are :
{code:java}
job-cluster --job-classname <classname> --fromSavepoint <savepointPath> --allowNonRestoredState
does not pick the savepoint path and does not start from it

job-cluster --fromSavepoint  <savepointPath>  -n --job-classname <classname>
-p works for savepoint retrieval{code}
 

  was:
Hello guys,

I've discovered that when deploying the job cluster on Kubernetes, the Job Cluster Manager
seems sensitive about the placement of arguments. 

For instance if i put the savepoint argument last, it never reads it. 

For instance if arguments are :

job-cluster --job-classname <classname> --fromSavepoint <savepointPath> --allowNonRestoredState
does not pick the savepoint path and does not start from it

job-cluster --fromSavepoint  <savepointPath>  -n --  job-classname <classname>
-p works for savepoint retrieval


> Flink on Kubernetes sensitive about arguments placement
> -------------------------------------------------------
>
>                 Key: FLINK-11900
>                 URL: https://issues.apache.org/jira/browse/FLINK-11900
>             Project: Flink
>          Issue Type: Bug
>    Affects Versions: 1.7.2
>            Reporter:  Mario Georgiev
>            Priority: Major
>
> Hello guys,
> I've discovered that when deploying the job cluster on Kubernetes, the Job Cluster Manager
seems sensitive about the placement of arguments. 
> For instance if i put the savepoint argument last, it never reads it. 
> For instance if arguments are :
> {code:java}
> job-cluster --job-classname <classname> --fromSavepoint <savepointPath> --allowNonRestoredState
does not pick the savepoint path and does not start from it
> job-cluster --fromSavepoint  <savepointPath>  -n --job-classname <classname>
-p works for savepoint retrieval{code}
>  



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message