Steps to reproduce:
- Run session.exec on a command that gets 'stuck' - doesn't complete - for me it was just 'bun install'
- The command never times out even with an explicit timeout.
- No commands work on that container again - the only solution seems to be to delete the container and recreate.
I've tried versions 0.5.6/0.6.11/0.7.8 - same result.
Is creating a new session for each worker request the best practice?