[ 
https://issues.apache.org/jira/browse/FLINK-32004?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Tan Kim updated FLINK-32004:
----------------------------
    Description: 
I set up access to the flink Web UI via the ingress settings in flinkDeployment.

Since this is an AWS environment, the ingress class uses ALB, not NGINX.

When using aws alb with flinkDeployment ingress settings when JobManager is 
restarted via autoscaler, intermittently ingress creation fails.

I say intermittent because ingress is generated irregularly, regardless of 
whether scaling is successful or not.

 

This issue occurs when the initial deployment always succeeds in creating 
INGRESS and accessing the FLINK WEB UI, but when it is scaled up/down via the 
autoscaler, INGRESS is not created sometimes.

  was:
I set up access to the flink Web UI via the ingress settings in flinkDeployment.

Since this is an AWS environment, the ingress class uses ALB, not NGINX.

When using aws alb with flinkDeployment ingress settings when JobManager is 
restarted via autoscaler, intermittently ingress creation fails.

I say intermittent because ingress is generated irregularly, regardless of 
whether scaling is successful or not.

 

This issue occurs when the initial deployment always succeeds in creating 
INGRESS and accessing the FLINK WEB UI, but when it is scaled up/down via the 
autoscaler, INGRESS is not created.


> Intermittent ingress creation when JobManager is restarted via autoscaler
> -------------------------------------------------------------------------
>
>                 Key: FLINK-32004
>                 URL: https://issues.apache.org/jira/browse/FLINK-32004
>             Project: Flink
>          Issue Type: Bug
>          Components: Autoscaler, Kubernetes Operator
>            Reporter: Tan Kim
>            Priority: Major
>
> I set up access to the flink Web UI via the ingress settings in 
> flinkDeployment.
> Since this is an AWS environment, the ingress class uses ALB, not NGINX.
> When using aws alb with flinkDeployment ingress settings when JobManager is 
> restarted via autoscaler, intermittently ingress creation fails.
> I say intermittent because ingress is generated irregularly, regardless of 
> whether scaling is successful or not.
>  
> This issue occurs when the initial deployment always succeeds in creating 
> INGRESS and accessing the FLINK WEB UI, but when it is scaled up/down via the 
> autoscaler, INGRESS is not created sometimes.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to