Server Version#: 1.15.1.780-a27ffa5be
Player Version#: Orca’s Samsung App v2.005
For the past few months, I’ve had issues with music on my iOS device (Streaming to in-car Bluetooth). I’d be listening to music, and the next song plays and is completely silent. Nothing comes out. I go back to previous song, there’s music. Back to next song, nothing. This has happened to multiple songs, and closing the app and reopening will sometimes allow the song to be played properly. 2nd issue with music, is that the song could be half way done, and immeditately skip to the next track. Doesn’t happen to every song, but have found that some songs do cause this more frequently.
For the past week, I’ve seen issues where my Plex randomly drops (Service is still running, but can’t access it externally or locally). Throws a 503 Service Unavailable. That’s what caused me to upgrade to 1.15.1.780. It’s sad when you contemplate coding a service checker to verify Plex is up, and if it isn’t, restart the service.
Now, with 1.15.1.780, I cannot stream anything to my TV running Orca’s Samsung App v2.005. This was working prior to this version. I don’t know what has changed, but this is not working anymore.
I need help on this one. I’ve been a Plex Pass Lifetime user for 5+ years, and haven’t had this many issues ever.
Also, is there a place where we can download older beta versions of PMS? I see no links to these older versions. I wanted to watch a show tonight, but if I can’t revert this change, I’m stuck unless I do a full reinstall or install a competitor.
Crashed on me 2 times today, listening to music. I just now upgraded to .791 and see that fixed a music transcoding issue on Windows (even though all of my music is Direct Play). We’ll see how that works and I’ll report back.
Just tried the Samsung TV again, and am now able to play TV Shows (.mkv) after installing this update as well.
Please try to separate the issues and also when reporting specifics with playing specific media, give the time and try to get logs from Plex client app as well
There was an issue with Samsung which could result in server locking out after few days and 503 responses would start to be returned. This was addressed in 1.15.1.780
(HTTP) Server could become unavailable when using the Samsung app (#9784)
The logs you provided show loads of frequent requests on the server PC - all coming through 127.0.0.1 - what Plex app did you have running on the server PC ? Was it Plex Web ? and which screen was it on?
The server was getting a request every 3-4 milliseconds which is excessive and i would like to understand how it arose. Was it a specfic screen you were on? Were there multiple Plex Web tabs open?
Example of the requests coming in every 3-4 ms
Mar 12, 2019 18:29:27.302 [14056] DEBUG - Request: [127.0.0.1:63838 (Loopback)] GET /player/proxy/poll?deviceClass=pc&protocolVersion=1&protocolCapabilities=timeline%2Cplayback%2Cnavigation%2Cmirror%2Cplayqueues&timeout=1 (12 live) GZIP Signed-in Token (Chris21788)
Mar 12, 2019 18:29:27.305 [7956] DEBUG - Request: [127.0.0.1:63193 (Loopback)] GET /player/proxy/poll?deviceClass=pc&protocolVersion=1&protocolCapabilities=timeline%2Cplayback%2Cnavigation%2Cmirror%2Cplayqueues&timeout=1 (12 live) GZIP Signed-in Token (Chris21788)
Mar 12, 2019 18:29:27.309 [1744] DEBUG - Request: [127.0.0.1:63838 (Loopback)] GET /player/proxy/poll?deviceClass=pc&protocolVersion=1&protocolCapabilities=timeline%2Cplayback%2Cnavigation%2Cmirror%2Cplayqueues&timeout=1 (12 live) GZIP Signed-in Token (Chris21788)
Mar 12, 2019 18:29:27.314 [13296] DEBUG - Request: [127.0.0.1:63193 (Loopback)] GET /player/proxy/poll?deviceClass=pc&protocolVersion=1&protocolCapabilities=timeline%2Cplayback%2Cnavigation%2Cmirror%2Cplayqueues&timeout=1 (12 live) GZIP Signed-in Token (Chris21788)
Mar 12, 2019 18:29:27.318 [12360] DEBUG - Request: [127.0.0.1:63838 (Loopback)] GET /player/proxy/poll?deviceClass=pc&protocolVersion=1&protocolCapabilities=timeline%2Cplayback%2Cnavigation%2Cmirror%2Cplayqueues&timeout=1 (12 live) GZIP Signed-in Token (Chris21788)
Mar 12, 2019 18:29:27.322 [9300] DEBUG - Request: [127.0.0.1:63193 (Loopback)] GET /player/proxy/poll?deviceClass=pc&protocolVersion=1&protocolCapabilities=timeline%2Cplayback%2Cnavigation%2Cmirror%2Cplayqueues&timeout=1 (12 live) GZIP Signed-in Token (Chris21788)
Mar 12, 2019 18:29:27.327 [2920] DEBUG - Request: [127.0.0.1:63838 (Loopback)] GET /player/proxy/poll?deviceClass=pc&protocolVersion=1&protocolCapabilities=timeline%2Cplayback%2Cnavigation%2Cmirror%2Cplayqueues&timeout=1 (12 live) GZIP Signed-in Token (Chris21788)
Mar 12, 2019 18:29:27.331 [13036] DEBUG - Request: [127.0.0.1:63193 (Loopback)] GET /player/proxy/poll?deviceClass=pc&protocolVersion=1&protocolCapabilities=timeline%2Cplayback%2Cnavigation%2Cmirror%2Cplayqueues&timeout=1 (12 live) GZIP Signed-in Token (Chris21788)
Mar 12, 2019 18:29:27.334 [7956] DEBUG - Request: [127.0.0.1:63838 (Loopback)] GET /player/proxy/poll?deviceClass=pc&protocolVersion=1&protocolCapabilities=timeline%2Cplayback%2Cnavigation%2Cmirror%2Cplayqueues&timeout=1 (12 live) GZIP Signed-in Token (Chris21788)
Mar 12, 2019 18:29:27.338 [14056] DEBUG - Request: [127.0.0.1:63193 (Loopback)] GET /player/proxy/poll?deviceClass=pc&protocolVersion=1&protocolCapabilities=timeline%2Cplayback%2Cnavigation%2Cmirror%2Cplayqueues&timeout=1 (12 live) GZIP Signed-in Token (Chris21788)
If you enable verbose logging on the server, we can get to find out more and also what version of the app
Settings / Server / General / Show Advanced
Set both verbose + debug
restart the server and try and recreate what was running then
and capture the logs
@sa2000, thanks for the reply. As for the music issue, it’s been working much better lately, but today (on my 2 hour drive), I had a few songs play a few seconds and skip to the next song. I didn’t grab times, but next time it happens, I’ll note it and create a new thread.
Samsung TV playing is fixed after I upgraded to 791.
So for now, let’s use this thread to focus on Plex becoming unavailable. I had it crash today (wasn’t streaming/watching anything), and tried to bring Plex up around 7:15pm ET.
Looking through the logs, I see the first error in the Plex Scanner log:
Mar 16, 2019 18:57:58.799 [8364] ERROR - Error issuing curl_easy_perform(handle): 28
Mar 16, 2019 18:57:58.800 [8364] DEBUG - HTTP simulating 408 after curl timeout
Mar 16, 2019 18:57:58.800 [8364] DEBUG - HTTP requesting GET http://127.0.0.1:32400/:/metadata/updateProgressMessage?message=Scanning%20Flushed%20Away%20(2006)
I unfortunately didn’t have verbose logging on. I’m assuming that log rotation will prevent these from blowing out my storage…It’s been roughly 3 days of running before this crashed again. (Also assuming that personal info is blocked out? eg tokens, etc).
I’ve enabled verbose logging and will report back when I crashed again.
Thank you for the logs. I can see what looks like a lockout that started when requests were getting stuck from around 07:44 am onwards - the requests being stuck were relating to music transcoding through a WAN connection.
To investigate would need to treat this as a deadlock and try and gather similar diagnostics as we normally do for a deadlock
So on next occurrence we need to get the connections list at the time and a process dump to go with the server logs.
To get the connections, you will need to know the server security token string as that would need to be added to the request that you would make in the browser when the issue arises.
When you get the lockout as happened here, load browser on the server and go to this url http://127.0.0.1:32400/connections?X-Plex-Token=xxxxxxxxxxxxxxxxx substituting the server security token for the xxxxxxxxxxxxxxxxx
This will display the list of connections. Select all the displayed text and copy into a text file. Note down the time
Next is to force a process dump. Use the 32-bit version of windows task manager - opening C:\Windows\syswow64\taskmgr.exe with right click and selecting Run As Adminstrator and then locate and select the Plex Media Server.exe process and right click and select Create Dump File
Then restart and download the logs.
Zip the dmp file and the saved connections list text file and upload to dropbox or similar service with the logs zip and send me link by Private Message.
With regards to the frequent requests i highlighted before - one every 3-4 milliseconds, the cause has been identified and it arises at the moment when you have multiple Plex Web browser tabs open. Restrict to one tab per browser for now.
@sa2000, had another crash this morning. So after yesterday, I did clear out all logs on my box, so the link I’ll be sending you through PM should have logs since then. Crash happened at 8:04am ET. Connection file was pulled at 8:27am ET.
