Post deployment checks missing

[itamararjuan]

When nebula finishes running we should check the health of the systems which were just provisioned and make sure it's alive and syncing correctly, this issue is here to document the brainstorming of how to go about doing that.

Couple of initial issues/questions I can think of and that we should discuss

• Metrics fetching from the node (SSH/HTTP)
• Which metrics consitute a healthy state
• Healthcheck route on public api of the node (when/eta/owner)
• How to validate the ethereum endpoint provided is in sync and available? (Pre nebula create and also after the node has been setup)
• Should we poll these metrics from Nebula while running the create command?
• Should we block the terminal and wait for these checks to show good healthy signs before we give back control to the user?

These are my initial thoughts and contemplations on post-deployment checks

Would love @electricmonk , @netoneko @erankirsh to pitch in and give their opinion.

[netoneko]

I believe that generally it's a waste of time. We can use existing network-status check that polls the metrics to understand if the network advances or the node is syncing, but I don't see anything on top of that as valuable.

And obviously no SSH unless we absolutely have to (and we don't have to because all the information about the chains is exported through metrics).