Hello, I have implemented a Kafka ES Connector by using the Kafka Connect Java API. The java application is deployed as a docker container using Marathon framework. There is no health check set up as of now.
The Marathon app fails every 2-3 minutes. Mesos Slave logs show the following message repeatedly whenever the task fails: E0509 11:40:11.918833 1231 slave.cpp:3252] Failed to update resources for > container de5c1573-3fa3-4b3e-a560-de6491cf3c28 of executor > 'infra_connect-kafka-es.0d1b37c4-15da-11e6-9215-02424ea4e672' running task > infra_connect-kafka-es.0d1b37c4-15da-11e6-9215-02424ea4e672 on status > update for terminal task, destroying container: Failed to determine cgroup > for the 'cpu' subsystem: Failed to read /proc/20764/cgroup: Failed to open > file '/proc/20764/cgroup': No such file or directory I ran the Docker container on each slave without using Marathon and there was no failure. I have checked out the issue ( https://issues.apache.org/jira/browse/MESOS-1837) but the comments are not so clear about a specific workflow to be adopted to fix this. Did anyone else have a similar issue? If yes, please suggest me a workaround for the same. Thanking you -- Anish Samir Mashankar
