Server Version#: 1.31.1.6733
Player Version#: Apple TV 8.15
For the past week, I’ve had many of my remote users, and myself, experiencing outages when streaming content from my server. Users are all direct playing either h264 or x265 encoded content. When I start something, it plays for 5-10 minutes, but eventually throws an error similar to “Can’t continue playback because the server is not responding”. The problem has gotten worse over the past few days, extending to more and more users, but it is mostly the ones who use an Apple TV that are experiencing the most strife. I run a variety of services on my unRAID system (PMS, PMM, automation apps, Tdarr, a few others) but my CPU is sleeping, hardware accelerated transcoding is working as expected and my memory usage is well within normal ranges. I cannot find a smoking gun here to determine what is going on, other than chalking it up to a corrupt or bad database and considering starting from scratch on the system. I’d rather not do that, and hope that someone on the Plex team can review my logs to see what is going on. I can see some Webhook warnings and a few errors corresponding with times when some of my users received the error, but I don’t know the fix. Any and all help would be greatly appreciated, thanks.
One other detail: I had found a random fix on a Reddit post where a user suggested that I shut down the docker, move my db and blobs.db files into another directory, start up the Plex docker and allow it to create new db/blobs.db, then shut Plex down again, delete the new files, move the old ones into the directory and start things up. I did do this, and noticed an immediate increase in speed and responsiveness when using the Plex in apps that are working (browser, iOS, PS5, etc). I don’t know what this is doing or why it “works”, but Apple TV users are still experiencing the same issue.
I changed my DB cache size from 40mb to 2000mb and started a series of local streams from ATV, Xbox, iOS and Safari browser. I also have 3 remote users with ATVs running content. There has not been a single stoppage yet. I want to wait until tomorrow, or maybe even later in the week to further verify this has remediated my issue, but things seem to be operating normally now. I don’t understand how this might have solved the problem, but I’m marking this now for future reference. I’ll update later today, or later this week, with further findings.
