ping-health

[snadrus]

Fixes #851
Caches the relevant alert results for the ping.

Refreshes every 5 minutes.

[snadrus]

@beck-8

[LexLuthr]

Approach is what I wanted to do but 30 minutes window is too big practically for the use case. We probably want to have a much lower window like 5 minutes for these 3 checks. Should we have one more task with PDP specific checks? We don't want to keep saying we are unhealthy for 30 minutes when there was a transient sync issue.

AJ: 👍

[snadrus]

@LexLuthr what's a PDP-specific check?
We could check if PDP tasks are seeing "lots of failures"

[LexLuthr]

Looks good. Can we add alert timing tests?

[LexLuthr]

We should probably update the docs to new timing changes. It is no longer every 1 hour but every top of the hour.

[snadrus]

Looks good. Can we add alert timing tests?

what is that?