Description
i used marathon+mesos to launch my apps.
when i update my app's image configuration and deploy it, i found a task failed, and then a new task is launched. Meanwhile, i use "docker ps" to get all containers, i found the failed task was launched, and the status of this task in mesos is failed and completed.
mesos lost the management for this failed task.
is this a bug?
steps to reproduce:
1. update image configuration, and redploy
2. make sure image be pulled after a long time
logs for this:
State
TASK_FAILED
Message
Failed to launch container: Failed to run 'docker -H unix:///var/run/docker.sock inspect mesos-f82fd98b-ba99-4ffe-bab5-eaa9dab8d5cc-S1.d84d9d59-218d-47fd-b930-e0d2359e61fa': exited with status 1; stderr='Error: No such image, container or task: mesos-f82fd98b-ba99-4ffe-bab5-eaa9dab8d5cc-S1.d84d9d59-218d-47fd-b930-e0d2359e61fa '
docker ps | grep mesos-f82fd98b-ba99-4ffe-bab5-eaa9dab8d5cc-S1.d84d9d59-218d-47fd-b930-e0d2359e61fa