Feature request: add active probing of GH systems

[nkaretnikov]

IIUC, the current approach is just to parse the atom feed, which is fetched periodically.
Would be cool to instead probe various GH sub-systems in case the official page is not reliable.

[mrshu]

@nkaretnikov interesting idea! Do you happen to have more thoughts on how that would work?

[nkaretnikov]

@mrshu I don't think it would be possible to do by keeping infra and hosting on GH.

So you'd need outside hosting and infra and then you'd define certain actions which would be representative of various subsystems that would run periodically, say, every 30 mins. If an action fails to run or there's no communication after N s, treat it as a failure and record it in the dashboard. Basically a keep alive check for pushes to the repo, creating PRs, running GH actions, etc. All that you can script using the GH API.

So this is a bit more work compared to what you have right now, not sure if you want to head in that direction.