Monitor: Best solutions for keeping track of uptime?

סיכום מבוסס בינה מלאכותית

The discussion revolves around monitoring the uptime and performance of a Discourse site, with users seeking recommendations for tools. hunterowens asks about monitoring DB, site, and Redis uptime, as well as response times, mentioning past use of New Relic. riking points to an existing New Relic plugin for Discourse. justin_gordon shares his experience using UptimeRobot and emphasizes the importance of maintenance, noting how easy it is to keep a Discourse site running without realizing issues. sam responds that while their recommendations differ due to factors like API, IPv6 support, and long-term reporting, they currently use Pingdom and are migrating to StatusCake due to IPv6 issues at scale.

Does anybody know or has written anything about monitoring a discourse site:

Ie, - DB uptime/Site Uptime/Redis Uptime
response times, etc.

I’ve used New Relic in the past but have no strong preferences.

There is already a New Relic plugin for Discourse:

I just signed up for https://uptimerobot.com/.

I’d like to get the latest recommendations from the Discourse dev team. @sam?

I thought about this today as I discovered that my programming site https://forum.shakacode.com was not responding. Yikes! Logging into the server and restarting the container

This is all I needed to get the site back up:

cd /var/discourse/
./launcher restart app

I then did the upgrade:

git pull
./launcher rebuild app

One “problem” with discourse is that it’s so easy to keep running that it’s easy to forget how to do maintenance!

Our recommendations are a bit tricky here, we decide on a tool based off a very different decision matrix to what a self-hoster would. We think about things such as API, clean IPv6 support, long term reporting and so on.

As it stands we use https://www.pingdom.com and are in the process of migrating to https://www.statuscake.com/ due to some very specific ipv6 issues we are seeing at pingdom at scale.