Do you know if your current buffering 33% are still due to the transcoder crashing ?
The 33% buffering text can be displayed for different underlying reasons - basically the roku app not receiving the transcode segment that it is asking for
Thank you for your assistance @sa2000 it is appreciated.
The last report I endured did demonstrate the same fault that has been occurring the previous two times. Please let me know if I have this misunderstood.
At the very least, it is quite impressive that Plex still continues to display a spinning 33% indicator to the user rather than a simple message with a code to help the user know what the heck is going on with your application and to (more importantly) report it back to you.
(It should be reporting it to you automatically, but I digress)
If a Plex Transcoder dmp file is generated on the server at the time of the 33% message then it is due to the transcoder crashing
and you could periodically get me server logs and the dmp file just for me to confirm if it is still crashing within the NVIDIA decoder module
If a Plex Transcoder dmp file is generated on the server at the time of the 33% message then it is due to the transcoder crashing
I believe I provided this in the last batch of artifacts that were sent to NVidia, please confirm.
and you could periodically get me server logs and the dmp file just for me to confirm if it is still crashing within the NVIDIA decoder module
I was under the impression that NVidia is aware of the problem and wanted another capture to re-re-confirm this issue in November 2022 as a third capture. Have I misunderstood what is occurring here?
what i was saying here is that 33% could arise for different reasons - one of which being the transcoder crashing. If a dmp file is created at the time of the 33% or just before then it is due to a transcoder crash
yes the crashes i saw in the NVIDIA module were the ones referred to NVIDIA
I was just suggesting sending new ones when they arise so i can analyze and see if the crash is still as before. Not all Transcoder crashes relate to the NVIDIA decoder
I was just suggesting sending new ones when they arise so i can analyze and see if the crash is still as before. Not all Transcoder crashes relate to the NVIDIA decoder
Thank you for this, I will keep this in mind. I appreciate your continued dialogue and assistance in working with me. It has abated my nasty nags on Twitter for the moment, at least. 
However, I feel I would be doing a disservice to both of us without pointing out that it really seems like the process around this would have been improved by now. It is going on two years now that we’ve been having to do this very arduous process for something that seems like should be automated and reported via your own telemetry.
In my own application development, every error is captured and reported to me. I am very aware of where errors occur and how often. I run a web-hosted application, which is of course, different from an os-hosted application such as Plex. However, I could not possibly see creating an application that is scaled across so many installations such as Plex without having the most basic reporting instrumentation and instead forcing that on the role of the customer to get this data to me.
Have you brought this up internally with the management there, by chance? Are there efforts underway to improve this process? Or has it already been culturally codified in Plex that making your customers full-time quality assurance engineers is an acceptable policy?
Thank you. I am following this up.
1 Like
I let it run this morning and hit something about 30 minutes in.
Faulting application name: Plex Transcoder.exe, version: 0.0.0.0, time stamp: 0x6322dd59
Faulting module name: ntdll.dll, version: 10.0.19041.2130, time stamp: 0xb5ced1c6
Exception code: 0xc0000374
Fault offset: 0x00000000000ff6a9
Faulting process id: 0xaf0
Faulting application start time: 0x01d95d8e5346d22f
Faulting application path: C:\Program Files\Plex\Plex Media Server\Plex Transcoder.exe
Faulting module path: C:\WINDOWS\SYSTEM32\ntdll.dll
Report Id: 6cc79bf5-64b7-43b5-8e7c-8914a9099fc1
Faulting package full name:
Faulting package-relative application ID:
Unfortunately, I do not see a crash dump and this may be due to having to install x64 plex due to facing this issue: Plex Media Server 1.30.0.6359 - New Transcoder Preview - #61 by Hossy923
Now it appears I have a weird install where I have a Plex Media Server installed on my C: and also my media drive.
Additionally, the LocalDumps as described in this article was missing so I re-added it.
Doesn’t take too long
Faulting application name: Plex Transcoder.exe, version: 0.0.0.0, time stamp: 0x6322dd59
Faulting module name: nvcuvid.dll, version: 7.17.14.5751, time stamp: 0x5fba2e65
Exception code: 0xc0000005
Fault offset: 0x00000000001e71bc
Faulting process id: 0x3be4
Faulting application start time: 0x01d95db40f5b7ae5
Faulting application path: C:\Program Files\Plex\Plex Media Server\Plex Transcoder.exe
Faulting module path: C:\WINDOWS\SYSTEM32\nvcuvid.dll
Report Id: 74833507-df23-410e-836e-65caf9688067
Faulting package full name:
Faulting package-relative application ID:
Ok - so the event log can tell us if it is still crashing in the Nvidia DLL
1 Like
Less than an hour to reproduce consistently. Has an update been provided to NVidia? I am unclear on the status of this issue. Thank you for any clarification/update you can provide.
No feedback from nVidia
I suggest you try to report the Faulting module path: C:\WINDOWS\SYSTEM32\nvcuvid.dll to nVIDIA and provide the dumps and driver details to them and see if they would investigate it - and they can reach out to us if they need any help from our end
I suggest you try to report the
Faulting module path: C:\WINDOWS\SYSTEM32\nvcuvid.dllto nVIDIA
Pardon my confusion here, but I thought that is what you have done now several times already.
Why are you suggesting this now rather than 2 years ago when we determined it was a fault occurring with NVdia?
I am very aware I could report this to NVidia but thought you were able to leverage your existing connection with them to get this escalated. I would appreciate further clarification here.
We have no direct contact with the NVIDIA Desktop team - we did report it to NVIDIA through our NVIDIA SHIELD contact who passed it on to their desktop team. We did ask for an update. I will ask again - it is not a direct route to the NVIDIA Desktop team
All I am suggesting is that coming to this through the other angle may help - as a user you are getting crashes reported in their dll file logged in the windows event log - so you can report it directly to NVIDIA,
1 Like
As a user, it sounded like I had a more direct/valuable connection so I was being lazy and letting that work out. Thank you for the further context. It is disappointing, of course, to see something take so long, but I suppose I should not be too surprised here.
I have opened a support ticket on NVidia’s side, @sa2000. If you would like to point your internal contact towards it, I would appreciate it. Ticket #230405-000032.
Thank you. 
Thank You. Will try and get that information to NVIDIA to link their two investigations
1 Like
Sooooo interesting turn of events here. I had 1.29 installed due to the pixelated live tv issue which was keeping me from upgrading.
I upgraded to 1.31 which (turns out) does not have the above problem and have been running streams on two Roku devices now for 2 days, both going on for more than 8 hours without issue. I do not think I have seen this before. It’s gotten to the point where I am wanting to mention something here to tempt fate in knocking me down a peg or two in my expectations here. 
This would be really something if 1.31 fixed both the pixelated Live TV issue AND this issue. I will continue to run this and update here. 
