Server Version#: 1.41.6.9685
Player Version#: 4.145.1
Transcoding appears to start and then fail to generate additional chunks on the latest server version when using a browser. Chunks continue to be generate and function properly when triggering transcoding via an app (even the latest broken Android app).
Anyone else experiencing this?
Cannot replicate.
can you please:
Attached.
Plex Media Server Logs_2025-04-17_23-10-18.zip (2.0 MB)
web_client_debug_log.txt (126.8 KB)
thanks for the logs. I found the issue immediately.
DNS failure.
Apr 17, 2025 21:09:52.499 [140039051189048] WARN - [HttpClient/HCl#2] HTTP error requesting GET https://plex.tv/api/v2/server/access_tokens?auth_token=xxxxxxxxxxxxxxxxxxxx (6, Couldn't resolve host name) (Could not resolve host: plex.tv)
Apr 17, 2025 21:09:53.666 [140039051189048] WARN - [HttpClient/HCl#4] HTTP error requesting GET https://plex.tv/api/v2/user?includeSubscriptions=1&includeProviders=1 (6, Couldn't resolve host name) (Could not resolve host: plex.tv)
Apr 17, 2025 21:09:54.064 [140039051189048] WARN - [HttpClient/HCl#5] HTTP error requesting GET https://plex.tv/api/v2/release_channels?X-Plex-Token=xxxxxxxxxxxxxxxxxxxx (6, Couldn't resolve host name) (Could not resolve host: plex.tv)
Apr 17, 2025 21:09:54.086 [140039051189048] WARN - [HttpClient/HCl#7] HTTP error requesting GET https://plex.tv/api/v2/user/privacy?X-Plex-Token=xxxxxxxxxxxxxxxxxxxx (6, Couldn't resolve host name) (Could not resolve host: plex.tv)
Apr 17, 2025 21:09:54.111 [140039051189048] WARN - [HttpClient/HCl#8] HTTP error requesting GET https://plex.tv/api/v2/features?X-Plex-Token=xxxxxxxxxxxxxxxxxxxx (6, Couldn't resolve host name) (Could not resolve host: plex.tv)
Apr 17, 2025 21:09:56.954 [140039051189048] WARN - [HttpClient/HCl#b] HTTP error requesting GET https://plex.tv/media/providers?X-Plex-Token=xxxxxxxxxxxxxxxxxxxx (6, Couldn't resolve host name) (Could not resolve host: plex.tv)
Also, you are using a FQDN which PMS does not know. plex.e******.com
Your certificate (P12) must include the Key, Cert, and CA
(PMS verifies the Cert with the Key and CA)
(HTTPS isn’t going to work with this happening)
Interesting. I did forget to put the key file in the correct location, which I just fixed. Thank you for catching that!
However, I did some DNS testing from the server in question as well as the client I used and things seem just fine for resolution… ping on the other hand seems to be responding for app.plex.tv but not for plex.tv. I’m assuming ICMP is just dropped for plex.tv but allowed for app.plex.tv
Server
root@hw-plex:/# dig plex.tv
; <<>> DiG 9.18.30-0ubuntu0.22.04.2-Ubuntu <<>> plex.tv
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 57388
;; flags: qr rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 0, ADDITIONAL: 1
;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 65494
;; QUESTION SECTION:
;plex.tv. IN A
;; ANSWER SECTION:
plex.tv. 40 IN A 34.253.120.192
plex.tv. 40 IN A 54.247.6.33
;; Query time: 3 msec
;; SERVER: 127.0.0.53#53(127.0.0.53) (UDP)
;; WHEN: Fri Apr 18 02:02:12 UTC 2025
;; MSG SIZE rcvd: 68
root@hw-plex:/# ping plex.tv
PING plex.tv (34.253.120.192) 56(84) bytes of data.
^C
--- plex.tv ping statistics ---
5 packets transmitted, 0 received, 100% packet loss, time 4082ms
root@hw-plex:/# ping plex.tv
PING plex.tv (54.247.6.33) 56(84) bytes of data.
^C
--- plex.tv ping statistics ---
2 packets transmitted, 0 received, 100% packet loss, time 1007ms
root@hw-plex:/# ping app.plex.tv
PING app.plex.tv (104.18.36.51) 56(84) bytes of data.
64 bytes from 104.18.36.51 (104.18.36.51): icmp_seq=1 ttl=57 time=2.39 ms
64 bytes from 104.18.36.51 (104.18.36.51): icmp_seq=2 ttl=57 time=2.19 ms
^C
--- app.plex.tv ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1002ms
Client Machine
PS C:\Users\jclif> nslookup plex.tv
Server: fw01-v100.redacted.com
Address: 10.100.0.1
Non-authoritative answer:
Name: plex.tv
Addresses: 54.247.6.33
34.253.120.192
Here’s what you should see for the nslookup calls.
[chuck@glockner tv.2013]$ nslookup app.plex.tv
Server: 127.0.0.53
Address: 127.0.0.53#53
Non-authoritative answer:
Name: app.plex.tv
Address: 104.18.36.51
Name: app.plex.tv
Address: 172.64.151.205
Name: app.plex.tv
Address: 2606:4700:4400::ac40:97cd
Name: app.plex.tv
Address: 2606:4700:4400::6812:2433
[chuck@glockner tv.2014]$ nslookup plex.tv
Server: 127.0.0.53
Address: 127.0.0.53#53
Non-authoritative answer:
Name: plex.tv
Address: 54.247.6.33
Name: plex.tv
Address: 34.253.120.192
[chuck@glockner tv.2015]$
Confirmed that’s what I see from the server:
root@hw-plex:~# nslookup app.plex.tv
Server: 127.0.0.53
Address: 127.0.0.53#53
Non-authoritative answer:
Name: app.plex.tv
Address: 172.64.151.205
Name: app.plex.tv
Address: 104.18.36.51
Name: app.plex.tv
Address: 2606:4700:4400::6812:2433
Name: app.plex.tv
Address: 2606:4700:4400::ac40:97cd
And confirmed from the client as well:
PS C:\Users\jclif> nslookup app.plex.tv
Server: fw01-v100.redacted.com
Address: 10.100.0.1
Non-authoritative answer:
Name: app.plex.tv
Addresses: 2606:4700:4400::6812:2433
2606:4700:4400::ac40:97cd
172.64.151.205
104.18.36.51
I’ve been looking at your logs in more detail and trying to figure this out.
I think I see:
Where it gets interesting .
(NVIDIA GP104GL [Quadro P5000])Apr 17, 2025 23:08:26.650 [140038875331384] DEBUG - [Req#300ca/Transcode] TPU: hardware transcoding: enabled, but no hardware decode accelerator found
Apr 17, 2025 23:08:26.650 [140038857386808] DEBUG - [Req#300ca/Transcode] Cleaning directory for session ggy7lue6j2xidc6ve0vgqosi (/opt/plextranscodes/Transcode/Sessions/plex-transcode-ggy7lue6j2xidc6ve0vgqosi-aaca76b0-a18e-45e6-b3eb-80e0febb254c)
Apr 17, 2025 23:08:26.651 [140038875331384] DEBUG - [Req#300ca/Transcode] [Universal] Using local file path instead of URL: /media/fs01/Plex Media/TV Shows/The Flash (2014)/Season 01/The Flash (2014) - S01E21 - Grodd Lives.mkv
Apr 17, 2025 23:08:26.651 [140038875331384] DEBUG - [Req#300ca/Transcode] TPU: hardware transcoding: final decoder: , final encoder:
I see the transcoder running but I don’t ever see HW engage.
Do you want the HW to be used? It looks like you do.
What happens when you type nvidia-smi at the command line?
Do you have packages libnvidia-decode-*** and libnvidia-encode-*** installed with the core vetted drivers ?
Why do you have multiple subnets? Multiple VLANs as well?
Thanks for digging in, it’s greatly appreciated!
Yes I have multiple VLANs for segmentation and isolation of specific items/services on my network. All sitting behind pfSense.
VLAN 100VLAN 200There are no issues connecting to the Plex server from any VLAN due to firewall rules. They have been functioning perfectly for about 5 years now.
Hardware transcoding works perfectly for everything, except a browser. Which is very odd. All streaming devices I’ve tested on that same VLAN 100 work perfectly with original quality or forcing a transcode, with the exception of browsers and the Windows Plex player.
Android on tablet, phone, and Nvidia Shield (2017) work flawlessly with transcoding. Remote users experience the same issue. All streaming devices work, browsers do not, unless the audio is the only item needing to be transcoded.
In testing from a browser, nvidia-smi does NOT show any transcode processes running like it does with the app based connections. This confirms what you were seeing regarding PMS deciding not to use the Quadro P5000.
As you can see below, the Quadro P5000 is indeed being used for transcoding for both decode and encode when using a Plex application instead of the browser. Confirmed by data being fed into Grafana from Telegraf as well – which I know pulls from this same source.
root@hw-plex:~# nvidia-smi
Fri Apr 18 16:05:51 2025
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.230.02 Driver Version: 535.230.02 CUDA Version: 12.2 |
|-----------------------------------------+----------------------+----------------------+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+======================+======================|
| 0 Quadro P5000 Off | 00000000:02:00.0 Off | Off |
| 38% 62C P0 55W / 180W | 149MiB / 16384MiB | 5% Default |
| | | N/A |
+-----------------------------------------+----------------------+----------------------+
+---------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=======================================================================================|
| 0 N/A N/A 3525 C ...lib/plexmediaserver/Plex Transcoder 147MiB |
+---------------------------------------------------------------------------------------+
Thanks for explaining .
I also use PfSense. (Love it)
Given past issues I’ve had with Chrome , would you consider:
info: I use Version 135.0.7049.95 (Official Build) (64-bit)
Done!
Client IP for testing was 10.100.0.104
Server IP is 10.254.0.42
Thanks for the updated log.
Apr 18, 2025 15:49:15.175 [140615204322104] WARN - [HttpClient/HCl#26] HTTP error requesting GET https://plex.tv/api/v2/server/access_tokens?auth_token=xxxxxxxxxxxxxxxxxxxx (6, Couldn’t resolve host name) (Could not resolve host: plex.tv)
DSM shell has nslookup and tracert. Time to start checking at the host level and tracking out to the internet.
This is a bare metal machine running Ubuntu 22.04.
File attached. Client is again 10.100.0.104
Plex Media Server Logs_2025-04-18_21-22-24.zip (5.6 MB)
I had a chance to run more nslookups and an mtr.
As previously, DNS resolution is working perfectly on the server.
Tracing to plex.tv, I did notice that somewhere along the line within TATA Communications (AS6453) (although Lumen on the device names) significant packet loss seems to be taking place.
app.plex.tv on the other hand takes a completely different path and has far less, but still significant packet loss.
root@hw-plex:~# nslookup plex.tv
Server: 127.0.0.53
Address: 127.0.0.53#53
Non-authoritative answer:
Name: plex.tv
Address: 34.253.120.192
Name: plex.tv
Address: 54.247.6.33
root@hw-plex:~# nslookup app.plex.tv
Server: 127.0.0.53
Address: 127.0.0.53#53
Non-authoritative answer:
Name: app.plex.tv
Address: 104.18.36.51
Name: app.plex.tv
Address: 172.64.151.205
Name: app.plex.tv
Address: 2606:4700:4400::ac40:97cd
Name: app.plex.tv
Address: 2606:4700:4400::6812:2433
root@hw-plex:~# mtr -w -c10 plex.tv
Start: 2025-04-18T20:36:51+0000
HOST: hw-plex Loss% Snt Last Avg Best Wrst StDev
1.|-- fw01-v200.redacted.com 0.0% 10 0.2 0.2 0.1 0.2 0.0
2.|-- tukw-dsl-gw69.tukw.qwest.net 0.0% 10 3.6 2.9 1.8 6.6 1.4
3.|-- 63-226-198-33.tukw.qwest.net 0.0% 10 3.0 9.8 2.4 44.3 13.7
4.|-- ??? 100.0 10 0.0 0.0 0.0 0.0 0.0
5.|-- 6453-3356-sea.sp.lumen.tech 0.0% 10 3.4 4.8 3.1 15.5 3.8
6.|-- if-ae-48-2.tcore1.00s-seattle.as6453.net 0.0% 10 144.8 145.3 144.7 146.6 0.7
7.|-- if-ae-20-2.tcore1.sv1-santaclara.as6453.net 0.0% 10 146.4 145.5 143.8 147.8 1.4
8.|-- ??? 100.0 10 0.0 0.0 0.0 0.0 0.0
9.|-- if-bundle-20-2.qcore1.sqn-sanjose.as6453.net 70.0% 10 145.0 145.0 144.9 145.1 0.1
10.|-- 63.243.218.16 60.0% 10 145.5 144.8 144.2 145.5 0.6
11.|-- 63.243.218.9 40.0% 10 144.6 144.6 143.8 145.4 0.6
12.|-- if-bundle-10-2.qcore2.ldn-london.as6453.net 10.0% 10 144.4 144.7 143.7 145.5 0.6
13.|-- if-ae-67-2.tcore2.ldn-london.as6453.net 0.0% 10 143.8 144.5 143.8 145.0 0.4
14.|-- 80.231.20.82 0.0% 10 145.2 151.1 144.0 179.7 13.0
15.|-- ??? 100.0 10 0.0 0.0 0.0 0.0 0.0
root@hw-plex:~# mtr -w -c10 app.plex.tv
Start: 2025-04-18T20:39:39+0000
HOST: hw-plex Loss% Snt Last Avg Best Wrst StDev
1.|-- fw01-v200.redacted.com 0.0% 10 0.1 0.2 0.1 0.2 0.0
2.|-- tukw-dsl-gw69.tukw.qwest.net 0.0% 10 2.9 4.0 1.8 16.2 4.3
3.|-- 63-226-198-33.tukw.qwest.net 0.0% 10 5.6 4.2 1.5 16.6 4.5
4.|-- ??? 100.0 10 0.0 0.0 0.0 0.0 0.0
5.|-- ae2.3615.edge6.Seattle1.net.lumen.tech 10.0% 10 14.5 5.8 2.3 14.5 4.3
6.|-- 4.30.140.62 70.0% 10 3.7 17.8 3.7 45.9 24.3
7.|-- 108.162.243.39 0.0% 10 4.5 4.4 3.6 8.0 1.3
8.|-- 104.18.36.51 0.0% 10 3.4 3.3 3.1 3.5 0.1
Out of curiosity, It appears (according to this history lookup site) that DNS entries were updated for plex.tv on 2025-03-25 which is fairly close to the time issues started cropping up for Plex scanners. Curious if there was a migration to a new back end hosting provider and things went a bit sideways:
I didn’t think of that. Good call.
Given PfSense can be very strict, which is why I use it,
Do you have any IP address-specific rules for Plex remote access ?
Asking because:
plex.direct to hairpin turn in PfSense (NAT Reflection required) – unless mitigatedprivate-domain: "plex.direct"
<MyPlex> in your logsAsking this because:
Simple test.
Have you, maybe I missed it, put the chrome machine on the same subnet as PMS – or made it completely remote ?
I do indeed have plex.direct entries for private-domain custom options for DNS resolver. I’ve tried using my cell phone on mobile only, with a browser and the exact same issue happens.
I’ve tested from the same client now on the server subnet 10.254.0.0 and the issue persists for just ranscoding.
In full transparency, this is sitting behind HAProxy which has worked flawlessly until a few weeks ago. Very odd that transcoding works in the applications but only fails in the browser. I don’t allow traffic on any port other than 443 inbound for Plex since it’s terminating at the HAProxy layer.
I have a secondary (service disabled) Plex server that is a standby in the HAProxy back end pool which gets an rsync copy of the plex folder structures using lsyncd. If the primary fails or I take it down for maintenance, HAProxy kicks the secondary into action. This has worked absolutely flawlessly for the past 4-5 years. More than happy to dig into any logs you’d like though!
As always, thanks for the assist here! Much appreciated.
I’ve been using Chrome for ages and started with PfSense in 2017.
My Chrome machine is on the LAN interface (workstation) with the server
TVs are on the DMZ interface (192.168.2.0/24)
Plex app on TVs is the only thing to cross into LAN
I don’t use any proxy or HA software. ( I’m not a public/commercial server )
Try this -
Purge / delete the Plex app/tokens from Chrome’s cache
Did one better. Win 10 VM on the same subnet as the Plex server. Never used Edge browser. Logged in, MFA’d, attempted to play.
Transcoding failed on every attempt with a message of: Playback Error
There was an unexpected error during playback. Please visit our forums if you continue to experience problems.
I then attempted playback from a new Chrome profile on the original laptop (vlan 100 – 10.100.0.0/24) and transcoding worked for the first minute as it has for every other test. I see the playback start, the progress/buffer indicator jump up and up and up and then stop. Playback continues fine until that buffer is exhausted and then it hangs.
Tautulli and nvidi-smi both confirm that the transcode did NOT use the Quadro P5000
I then attempted to use the new Plex Android app on my Samsung phone on the same network (VLAN 100) as the laptop. Direct play worked, then I forced a transcode which also worked flawlessly. Tautulli and nvidia-smi confirmed the use of the Quadro for transcoding.
This seems 100% isolated to use of the web client.
We just had a problem with chrome.
For reasons unknown, Chrome would not work (nothing would play), just how you report.
We jockeyed between chrome and firefox.
On first opening, firefox worked.
Close and reopen firefox - failed.
Reopened Chrome Plex/web – No access
As last change before purging chrome and starting over -
NO idea what got corrupted but that cleared it.
Please let me know your results.
Note: I did have to go to each of my test machines and sign in again but full functionality has been restored
