No Shell In Container Images

[infinitydon]

Hi,

I wanted to check if the removal of shell access in the container build was intentional. This makes troubleshooting a bit more difficult—though I understand a debug container can be injected. Would it be possible to provide the same user experience as when ellaCore is run directly as a system daemon, where shell access is still available for troubleshooting and status checks?

Thanks

[gruyaume]

Hello @infinitydon,

We intentionally only include what is absolutely necessary in the container image for Ella Core to function. This keeps the image lean (76.4MB at the moment) and more secure. Adding packages increases the attack surface and bloat the image's size.

For troubleshooting, pebble is the service manager in the container image and some pebble commands can be very useful for troubleshooting (ex. pebble exec <command>, pebble ls <dir>, ...). For status checks, hitting the HTTP API /status endpoint or executing pebble services ella-core can inform you about the service's health.

[infinitydon]

Thanks for the explanation, will checkout the pebble command