[
https://issues.apache.org/jira/browse/FLINK-30518?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Binh-Nguyen Tran updated FLINK-30518:
-------------------------------------
Description:
Since flink-conf.yaml is mounted as read-only configmap, the
/docker-entrypoint.sh script is not able to inject correct Pod IP to
`jobmanager.rpc.address`. This leads to same address (e.g flink.ns-ext) being
set for all Job Manager pods. This causes:
(1) flink-cluster-config-map always contains same address for all 3 component
leaders (see screenshot)
(2) Accessing Web UI when jobmanager.replicas > 1 is not possible with error
{code:java}
{"errors":["Service temporarily unavailable due to an ongoing leader election.
Please refresh."]} {code}
~ flinkdeployment.yaml ~
{code:java}
spec:
flinkConfiguration:
high-availability: kubernetes
high-availability.storageDir: "file:///opt/flink/storage"
...
jobManager:
replicas: 3
... {code}
was:
Since flink-conf.yaml is mounted as read-only configmap, the
/docker-entrypoint.sh script is not able to inject correct Pod IP to
`jobmanager.rpc.address`. This leads to same address (e.g flink.ns-ext) being
set for all Job Manager pods. This causes:
(1) flink-cluster-config-map always contains same address for all 3 component
leaders (see screenshot)
(2) Accessing Web UI when jobmanager.replicas > 1 is not possible with error
{code:java}
{"errors":["Service temporarily unavailable due to an ongoing leader election.
Please refresh."]} {code}
> [flink-operator] Kubernetes HA not working due to wrong jobmanager.rpc.address
> ------------------------------------------------------------------------------
>
> Key: FLINK-30518
> URL: https://issues.apache.org/jira/browse/FLINK-30518
> Project: Flink
> Issue Type: Bug
> Components: Kubernetes Operator
> Affects Versions: kubernetes-operator-1.3.0
> Reporter: Binh-Nguyen Tran
> Priority: Major
> Attachments: flink-configmap.png
>
>
> Since flink-conf.yaml is mounted as read-only configmap, the
> /docker-entrypoint.sh script is not able to inject correct Pod IP to
> `jobmanager.rpc.address`. This leads to same address (e.g flink.ns-ext) being
> set for all Job Manager pods. This causes:
> (1) flink-cluster-config-map always contains same address for all 3 component
> leaders (see screenshot)
> (2) Accessing Web UI when jobmanager.replicas > 1 is not possible with error
> {code:java}
> {"errors":["Service temporarily unavailable due to an ongoing leader
> election. Please refresh."]} {code}
>
> ~ flinkdeployment.yaml ~
> {code:java}
> spec:
> flinkConfiguration:
> high-availability: kubernetes
> high-availability.storageDir: "file:///opt/flink/storage"
> ...
> jobManager:
> replicas: 3
> ... {code}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
