Thanks in advance to anybody who can help. I rebuilt my Plex server recently. I’m using an ASRock DeskMini with an AMD Ryzen 5 3400G, 32GB of RAM, and a 1TB SSD. I’m using two HDD enclosures attached to the server via USB-C and USB 3.0 connections respectively. One of the enclosures was also used on my last machine, and the USB-C enclosure is a new 4 bay running all drives as single drives, so Raid is not a factor. I’m running into 3 major issues:
When I play media, Plex behaves inconsistently. Sometimes it plays the content nearly instantly, and other times it spins for long periods of time (10 seconds on the short end, with it indefinitely spinning on the long end). My old server with a much weaker processor played all media in about 3 seconds, so this behavior is alarming.
Once media is playing, it occasionally just freezes. When this happens, no video or audio is playing. It is almost like the media is paused but it’s not indicating any buffering is occurring. System resources on the host server are readily available (< 25% CPU usage and <20% RAM usage.). The same title may not react the same way twice either. For example it may play a movie from front to back with no problem, but on subsequent viewings the title may randomly freeze.
Similar to the second problem, but in these instances the title being played continues with the audio but the video is frozen. It requires the user to back out of the title and start playing it again to fix the issue.
I’m happy to do whatever work I need to so I can identify root cause. I just don’t know what to look for first since I’ve verified all of the hardware pieces appear to be working properly and there’s no shortage of resources on the host server that would be preventing transcoding.
Does this behaviour happen with any type of media (1080p, 4K, live tv etc. even if it’s inconsistent)? If you limit playback to direct steaming on your local network, do you have the same problems? When a client is direct streaming, and taking a quick look at task manager, are there any resources on your server that are over utilised?
The behavior happens across all media types, or at least I’ve not been able to see a trend. Most of my media is 1080p or 720p though
Problem persists. I’ve noticed the problem on my local network, but I’ve also been told it happens on remote devices.
Taek manager/resource monitor says all resources are readily available. The most the CPU ever gets to is about 25% and the RAM stays under 20%.
I’ve also tried to go into the HDD Enclosures and disable them from sleeping in case the issue was plex entering sloth mode, and the HDD’s not waking up before the buffer needed to restart. No luck. What’s concerning to me is even the initial plex menu screens seem to take longer to load and buffer when selecting a title (prior to initiating playback). I feel like the plex install is corrupted somehow or is just not performing well on this build, but I restored it from a previous backup exactly following directions so I’m a bit lost why it would perform worse on a much better machine.
I know what you mean; that’s really frustrating. What were the specs on your last machine? What version of PMS were you running?
I’ve seen people have good results from database repairs. The database might have issues after the restore from backup? See here.
I’ll keep going through your logs. Can you find a video that direct plays, reboot Plex, test, and export logs again? Might be worth trying to isolate to a single stream and then adding complexity.
Thanks so much again for helping out. I feel like I’m losing my mind.
Here are hopefully some logs that might help. I had three instances of movies freezing up tonight around 18:00, 18:43ish, and 18:51ish. I’ll attach the latest log pull from a few minutes after the last freeze to hopefully help narrow it down.
My previous server was on an older i7 920 intel processor with 16GB of DDR3 RAM, so it was pretty outdated. The new one being on an R5 3400G with 32GB of DDR4 2933 should be much more powerful with plentiful resources for at least 4 simultaneous 1080p transcodes according to the benchmarking thresholds. I’m also on a 300/50 fiber internet connection with an optimized tri-band wireless AC setup for in-home wireless, and all of my connectivity/speed tests are rock solid on the devices I’m streaming from. The Plex server is also hard wired to the router to help alleviate any potential problems there. The router is the same one that was used during the last build though so I’m pretty certain the networking side of things is not going to be in play as a factor.
My old server was always running the latest release build of Plex, so when I transferred over was around November-ish. Don’t recall the exact build but I made sure I used the same versions on both machines during the migration to mitigate any potential conflicts. (I’m also on the latest now… 1.18.6.2368)
If these logs don’t help I’m certainly willing to try to isolate a single stream to test, but I think these last 3 tonight happening so close together may produce a pretty good result. The first 2 were on the same movie, the third one on a different movie, but all 3 occurred on the same streaming device.
Is there any way to get a moderator or Plex employee to check into this? I provided multiple sets of logs and as many details as I can think of, but I’m still sitting here with no resolution or clear path forward over a week later.
I spent some time going through your logs, though it would be helpful if someone more knowledgeable could have a look.
Do you have a throughout limit on outbound connections? There’s a lot of transcoding happening for a 50Mb up connection. Is auto quality enabled? Try disabling that and hard setting something like 4Mbps even for watching locally.
