Maxence Cramet created AIRFLOW-6810:
---------------------------------------

             Summary: KubernetesPodOperator pod is completed but xcom side car 
is stuck
                 Key: AIRFLOW-6810
                 URL: https://issues.apache.org/jira/browse/AIRFLOW-6810
             Project: Apache Airflow
          Issue Type: Bug
          Components: executor-kubernetes
    Affects Versions: 1.10.6
            Reporter: Maxence Cramet
            Assignee: Daniel Imberman


We're using KubernetesPodOperator with param xcom_push=true in order to push 
information from our task.

>From time to time the main pod completes but the side car pod is stuck.

Here's the output of the pods details:
{noformat}
kubectl describe pod my_pod
Name:               my_pod
Namespace:          default
Priority:           0
PriorityClassName:  <none>
Node:               xxx
Start Time:         Wed, 05 Feb 2020 11:12:33 +0000
Labels:             xxx
Annotations:        xxx
Status:             Running
IP:                 xxx
Containers:
  base:
    Container ID:  xxx
    Image:         xxx
    Image ID:      xxx
    Port:          <none>
    Host Port:     <none>
    Args:
      xxx
    State:          Terminated
      Reason:       Completed
      Exit Code:    0
      Started:      Wed, 05 Feb 2020 11:12:38 +0000
      Finished:     Wed, 05 Feb 2020 11:12:47 +0000
    Ready:          False
    Restart Count:  0
    Limits:
      memory:  512Mi
    Requests:
      memory:  512Mi
    Environment:
      xxx
    Mounts:
      /airflow/xcom from xcom (rw)
  airflow-xcom-sidecar:
    Container ID:  
docker://83053d7d292cda9156454ac13064d72ace1e4f72738ba9b62b04ff57cb7966cc
    Image:         alpine
    Image ID:      
docker-pullable://alpine@sha256:ab00606a42621fb68f2ed6ad3c88be54397f981a7b70a79db3d1172b11c4367d
    Port:          <none>
    Host Port:     <none>
    Command:
      sh
      -c
      trap "exit 0" INT; while true; do sleep 30; done;
    State:          Running
      Started:      Wed, 05 Feb 2020 11:12:40 +0000
    Ready:          True
    Restart Count:  0
    Limits:
      memory:  4Gi
    Requests:
      cpu:        1m
      memory:     2Gi
    Environment:  <none>
    Mounts:
      /airflow/xcom from xcom (rw)
      xxx
Conditions:
  Type              Status
  Initialized       True 
  Ready             False 
  ContainersReady   False 
  PodScheduled      True 
Volumes:
  xcom:
    Type:       EmptyDir (a temporary directory that shares a pod's lifetime)
    Medium:     
    SizeLimit:  <unset>
  xxx
QoS Class:       Burstable
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:          <none>{noformat}
I don't have more information of the possible causes of that.

 



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

Reply via email to