Figure out why status monitor frequently reports ~1m downtime on services

[jasonappah]

Don't know what's going on with this, but the extra noise these notifications create in Discord is annoying and is starting to make me tune out #incidents, which is obviously bad if real downtime occurs. We need to figure out why this downtime gets reported and if it's a real issue, address it. If it's just an issue with our monitoring setup, take care of that as well. We recently switched from Better Uptime (managed monitoring service) to Uptime Kuma (self-hosted monitoring service running on a VPS we control) and wonder if it might be worth switching back.

[stationswift001]

Should be as simple as increasing the time between pings. I'll change it sometime today

[stationswift001]

Call it whatever you want ("real fix" "bandaid solution", whatever) but I upped the ping interval to be every 5 minutes, so hopefully we get more lucky with the random one minute downtimes. LMK if settings need to be adjusted further, iirc Uptime Kuma can check for a string on a page if you wanna get super specific about uptime

[jasonappah]

thanks!

[jasonappah]

it still does the thing :/

FWIW better uptime has been reporting 502s at least once per day for all services so this might be an actual problem to figure out

[democat3457]

Also, occasionally we do actually get 502s - last night, someone tried to access learn.utd.ms and it was actively server erroring. It seems to routinely get 502 every 3 hours... is something timed to restart every 3 hours?

[ericmutton]

Moodle may have an improper configuration considering just how many moving parts make it work. Is this a false positive or is Moodle experiencing service degradation?

[ericmutton]

Considering how exhaustive this reporting is, perhaps something between Kuma itself and the Webhook is in play.