I just have uptime service hosted outside of our main infra. It connects to my service called Siren, which alerts me on my phone with an alarm on full volume with SWAT cat intro.
It's good enough for what we do, barely have any downtime. But it helped me figure out 6s downtime we would experience when our spot instances get knocked out, so it helped me increase health check frequency
6s downtime is a lot when you are getting hammered at 100 RPS.
I actually don't.
I just have uptime service hosted outside of our main infra. It connects to my service called Siren, which alerts me on my phone with an alarm on full volume with SWAT cat intro.
It's good enough for what we do, barely have any downtime. But it helped me figure out 6s downtime we would experience when our spot instances get knocked out, so it helped me increase health check frequency
6s downtime is a lot when you are getting hammered at 100 RPS.