feat(ui): show reason for failure in endpoint events

[perfectra1n]

Summary

This PR shows the reason as to why an Endpoint failed, instead of just "it failed". E.g. right now you look at something like this:

and you ask yourself "But why did it fail? What criteria failed?"

So this is what I've designed at this point:

Checklist

• [ x ] Tested and/or added tests to validate that the changes work as intended, if applicable.
• [ x ] Updated documentation in README.md, if applicable.

[perfectra1n]

I'm not exactly sold on the look just yet, but I was curious to see what your opinion was, or what your appetite for such a thing would be. As a user I would like to know "why" and "what" failed with such a program, but you may have different intentions.

[TwiN]

This isn't a bad idea, but what if there are multiple reasons for the endpoint being unhealthy? For instance, what if, for the first 20 minutes, it's unhealthy because of the status, but from 20 to 40 minutes, it's unhealthy because of the response time?

[perfectra1n]

I'm sure that Health criteria can become quite complex - so if we have an example EndpointA with health criteria: criteria1, criteria2, and criteria3...

If it fails at first due to criteria1, then after 10 minutes (or some other time-based criteria requirement) fail due to criteria2, I've seen some platforms have the UI reflect that e.g.:

"failed due to criteria1"
"10 minutes later"
"failed due to criteria2"

What do you think of something like that? Or instead, we could show just all criteria that failed initially (or after X amount of time) and leave it to the user's discretion :)

[TwiN]

Perhaps having it as a set (unique values) would be sufficient for now.

Just a heads up, the current implementation is missing a lot of bits and pieces, the biggest one being persistence.

[perfectra1n]

Perhaps something like this then, where we aggregate the unique failures and show the timestamp of failure?