Checking executor logs is time consuming because we have to determine the slave, enter the docker container, find the correct file and then read it.
Imagine we have launched Kafka and it keeps failing and restarting. We have a bunch FAILED tasks and log directories in the sandbox. The tasks have IDs look like this:
Checking executor logs is time consuming because we have to determine the slave, enter the docker container, find the correct file and then read it.
Imagine we have launched Kafka and it keeps failing and restarting. We have a bunch FAILED tasks and log directories in the sandbox. The tasks have IDs look like this:
kafka-mesos-scheduler.708e9047-b73c-11e5-a399-0242ac110005
where the part after the dot keeps changing for every new task.
Let's add a minimesos logs command that accepts part of a task ID like this:
Perhaps it should log stdout by default and stderr when adding a --stderr flag or vice versa.