Running a container in kubernetes is sometimes annoyingly difficult. Especially if the feedbackloop is long, when you only get a failing container notification.
Here is one I had recently:
OCI runtime exec failed: exec failed: (...) executable file not found in $PATH": unknown
I knew the container had the thing, so what on earth is going wrong?
In a manifest for kubernetes (deployments/statefulsets etc.) you must specify
a container (of course) but you can also supply a cmd and or args.
The container that I started already executed a command on start, but by supplying the command with cmd, the system could not find that part.
It took me 2 hours but finally it dawned on me: supply the things in args.
Now it just runs.
I’m publishing this as part of 100 Days To Offload. You can join in yourself by visiting https://100daystooffload.com, post - 15/100
Find other posts tagged #100DaysToOffload here