Chore: context-box restarts when it can't ping MongoDB

[JKBGIT1]

Current Behaviour

If context-box can't ping MongoDB as part of its health check it restarts.

Expected Behaviour

Do we want to handle this problem this way? Why should restarting the context-box fix the connection to another party?

Steps To Reproduce

I experienced this when I deployed Claudie on a cluster without the required images. The context-box image got pulled before the Mongo one, therefore the pod with the context-box container was up before the pod with Mongo.
However, you can replicate it even when you prevent Mongo from scheduling.

[JKBGIT1]

I think keeping the context-box not ready should be enough because the restart most probably doesn't fix the connection issue.