`docker ps` checking fails randomly about che

[TylerJewell]

So is this a good fail? I am wondering if we ignored this error and allowed the Che image to boot, would it be functional?

Wondering if we should improve how we test for docker readiness.

[freewind]

@TylerJewell Sometimes it's good fails, but sometimes not. But can't distinguish it from the output. For the time I posted this issue, I felt it more like a bug

[TylerJewell]

We might change this to do a full docker ps -a, which often will show active containers that are in a zombie state.

[freewind]

What about using docker info instead of docker ps?

If this check fails docker info will fail as well. The problem is that maybe docker does not have enough time to start while che.sh runs. Can you try adding sleep 10 to the CMD command?

I just started a fresh node on Digital Ocean - cannot reproduce the issue:

https://www.youtube.com/watch?v=UVucYKI5NHs

[freewind]

@eivantsov Thanks for the video. I will add such a sleep in the future if I have the same problem, and then will report it here

[freewind]

This issue seems not appear any more with latest version.

-v /var/run/docker.sock:/var/run/docker.sock should be the key when starting che. Thanks for fixing it