Random things you need to know to maintain a #PeerTube server for years on end:
✅ Upgrading PeerTube
✅ Changing Node versions
✅ Tweaking nginx configs from one version to the next
✅ Using systemctl and journalctl to isolate error messages as they happen
✅ Using Sentry to capture context around error messages
✅ Reading stack traces
✅ Checking directory permissions when something won’t upload
✅ Troubleshooting ffmpeg for transcoding
✅ Setting up a remote runner to offset video transcoding on a small server
✅ Dealing with Object Storage
✅ Infrastructure upgrades
✅ Restoring from backups
✅ Migrating data to a new server
✅ Using psql to parse the database and manipulate records
✅ Using redis-cli to examine and drop stuck jobs
✅ Searching for weird and obscure error messages
✅ Moderating and managing a community
✅ Taking donations
✅ Keeping everybody in the loop
Make no mistake: self-hosting is awesome, and being able to support a community is also awesome. It’s just that a lot of people get into hosting their own instances without realizing some of the crazy stuff they might have to do to keep things running.