When working with multiple containers running on a big cluster it's important to know what's where and how it performs. Monk provides simple but effective tools for inspecting and troubleshooting your workloads.
ps command will show you all containers currently running in your cluster at any time.
monk describe some/template
This will show a summary about a running workload, just like
monk run would, but without re-running it.
monk ps --allmonk ps -a
ps command with
-a for short) will list all containers, even the ones that were running but are currently not.
monk logs some/runnablemonk logs -c <container-id> some/runnable
logs command will show you the logs from any container, even the ones that are stopped. It's perfectly fine to just pass the runnable - monk will prompt show you a select if it detects more than one container.
monk logs --follow some/runnablemonk logs --follow -c <container-id> some/runnable
-f) option to stay get a continuous stream of logs from a running container.
monk shell some/runnablemonk shell -c <container-id> some/runnable
shell command opens an interactive shell to any running container. It's perfectly fine to just pass the runnable - monk will prompt show you a select if it detects more than one container.
exit command in the shell to end the session.
Will show the current CPU, memory, and disk usage of all containers running in the cluster.