Plex app not showing photos (metadata)

I’ve been searching everywhere regarding my issue but I can’t seem to find anything related to it, so I’m a bit lost on this kind of strange bug.
I’m running plex in a docker environment with Traefik as proxy, everything works fine in the web browser (photos, video, music, …) but when using the plex app it doesn’t show any photo (metadata), only the list and texts.
Playing video’s and music works but photos it won’t show, at a certain time I gave up but now after looking back in the logs I noticed a difference between the web and the app.
The plex app demands a photo to be transcoded but in it’s url request it references towards http://127.0.0.1:443. When using the web browser it’s without this referal to localhost.
I don’t understand why the app (on my phone, tablet, …) always uses localhost in the url request.
Any idea where I can find or further debug this referal to localhost?

Thanks.

Server Version#: Version 1.24.5.5173
Player Version#: 8.25.1.28703
Nov 20, 2021 18:10:20.501 [0xffff9a811aa8] DEBUG - Photo transcoder: Request for url [http://127.0.0.1:443/library/metadata/6293/thumb/1636842235] (is local: 0 upscaled: 0)
Nov 20, 2021 18:10:20.515 [0xffff9a811aa8] DEBUG - HTTP requesting GET http://127.0.0.1:443/library/metadata/6293/thumb/1636842235
Nov 20, 2021 18:10:20.516 [0xffff9a811aa8] ERROR - Error issuing curl_easy_perform(handle): 7
Nov 20, 2021 18:10:20.516 [0xffff9a811aa8] WARN - HTTP error requesting GET http://127.0.0.1:443/library/metadata/6293/thumb/1636842235 (7, Couldn’t connect to server) (Failed to connect to 127.0.0.1 port 443: Connection refused)
Nov 20, 2021 18:10:20.517 [0xffff9aa51aa8] DEBUG - Completed: [10.0.0.5:44400] 404 GET /photo/:/transcode?height=684&url=http%3A%2F%2F127.0.0.1%3A443%2Flibrary%2Fmetadata%2F6293%2Fthumb%2F1636842235&width=1080 (11 live) TLS GZIP 16ms 379 bytes (pipelined: 1)

1 Like

I have a similar problem with my thumbnails for movies/shows not showing up on my android devices(only happens on android, not with the web client or roku app). I have the same errors in my logs:

Nov 23, 2021 10:36:01.582 [0x7f6613f75b38] Debug — Request: [10.42.1.47:38252 (Subnet)] GET /photo/:/transcode?height=599&url=http%3A%2F%2F127.0.0.1%3A443%2Flibrary%2Fmetadata%2F7232%2Fart%2F1637633625&width=1080 (10 live) TLS GZIP Signed-in Token (durden0)
Nov 23, 2021 10:36:01.582 [0x7f6613f75b38] Debug — Photo transcoder: Request for url [http://127.0.0.1:443/library/metadata/7232/art/1637633625] (is local: 0 upscaled: 0)
Nov 23, 2021 10:36:01.583 [0x7f6613f75b38] Debug — HTTP requesting GET http://127.0.0.1:443/library/metadata/7232/art/1637633625
Nov 23, 2021 10:36:01.583 [0x7f6613f75b38] Error — Error issuing curl_easy_perform(handle): 7
Nov 23, 2021 10:36:01.583 [0x7f6613f75b38] Warning — HTTP error requesting GET http://127.0.0.1:443/library/metadata/7232/art/1637633625 (7, Couldn't connect to server) (Failed to connect to 127.0.0.1 port 443: Connection refused)
Nov 23, 2021 10:36:01.586 [0x7f6614684b38] Debug — Completed: [10.42.1.47:38252] 404 GET /photo/:/transcode?height=599&url=http%3A%2F%2F127.0.0.1%3A443%2Flibrary%2Fmetadata%2F7232%2Fart%2F1637633625&width=1080 (10 live) TLS GZIP 3ms 379 bytes (pipelined: 1)
Nov 23, 2021 10:36:01.589 [0x7f6613f75b38] Debug — Request: [10.42.1.47:38252 (Subnet)] GET /library/metadata/7232?asyncAugmentMetadata=1&checkFiles=1&includeExtras=1&includeReviews=1 (10 live) TLS GZIP Signed-in Token (durden0)
Nov 23, 2021 10:36:01.629 [0x7f6613f75b38] Debug — It took 10.000000 ms to retrieve 142 items.
Nov 23, 2021 10:36:01.640 [0x7f6613f75b38] Debug — It took 10.000000 ms to retrieve 142 items.
Nov 23, 2021 10:36:01.645 [0x7f6613f75b38] Debug — We're going to try to auto-select an audio stream for account 1.
Nov 23, 2021 10:36:01.645 [0x7f6613f75b38] Debug — Selecting best audio stream for part ID 24688 (autoselect: 0 language: en)
Nov 23, 2021 10:36:01.645 [0x7f6613f75b38] Debug — Audio Stream: 56906, Subtitle Stream: -1
Nov 23, 2021 10:36:01.652 [0x7f6613f75b38] Debug — Activity: registered new activity 6073854d-45ce-491d-8ff7-758d22f3e45c - "Refreshing"
Nov 23, 2021 10:36:01.653 [0x7f661440ab38] Debug — Activity: updated activity 6073854d-45ce-491d-8ff7-758d22f3e45c - completed 0.0% - Refreshing
Nov 23, 2021 10:36:01.653 [0x7f6613ca3b38] Debug — Augment: Completed movie augmentation in 0.0 sec.
Nov 23, 2021 10:36:01.653 [0x7f661440ab38] Debug — Activity: Ended activity 6073854d-45ce-491d-8ff7-758d22f3e45c.
Nov 23, 2021 10:36:01.655 [0x7f6614684b38] Debug — Completed: [10.42.1.47:38252] 200 GET /library/metadata/7232?asyncAugmentMetadata=1&checkFiles=1&includeExtras=1&includeReviews=1 (10 live) TLS GZIP 65ms 10077 bytes (pipelined: 2)
Nov 23, 2021 10:36:01.705 [0x7f6613f75b38] Debug — Request: [10.42.1.47:38252 (Subnet)] GET /hubs/metadata/7232/related?count=10&excludeElements=Actor%2CCollection%2CCountry%2CDirector%2CGenre%2CLabel%2CMood%2CPart%2CProducer%2CRole%2CSimilar%2CWriter%2CPhoto%2CVast%2CTopic&excludeFields=summary%2Ctagline%2Cfile&includeAugmentations=1&includeEmpty=1&includeExternalMetadata=1&includeLibraryPlaylists=1&includeMeta=1&includeRecentChannels=1&includeStations=1&includeTrailers=1&includeTypeFirst=1&libraryHubsOnly=1 (10 live) TLS GZIP Signed-in Token (durden0)
Nov 23, 2021 10:36:01.711 [0x7f66139f3b38] Debug — Request: [10.42.1.47:38258 (Subnet)] GET /photo/:/transcode?height=552&url=http%3A%2F%2F127.0.0.1%3A443%2Flibrary%2Fmetadata%2F7232%2Fthumb%2F1637633625&width=368 (11 live) TLS GZIP Signed-in Token (durden0)
Nov 23, 2021 10:36:01.712 [0x7f66139f3b38] Debug — Photo transcoder: Request for url [http://127.0.0.1:443/library/metadata/7232/thumb/1637633625] (is local: 0 upscaled: 0)
Nov 23, 2021 10:36:01.712 [0x7f66139f3b38] Debug — HTTP requesting GET http://127.0.0.1:443/library/metadata/7232/thumb/1637633625
Nov 23, 2021 10:36:01.713 [0x7f66139f3b38] Error — Error issuing curl_easy_perform(handle): 7
Nov 23, 2021 10:36:01.713 [0x7f66139f3b38] Warning — HTTP error requesting GET http://127.0.0.1:443/library/metadata/7232/thumb/1637633625 (7, Couldn't connect to server) (Failed to connect to 127.0.0.1 port 443: Connection refused)
Nov 23, 2021 10:36:01.713 [0x7f6614684b38] Debug — Completed: [10.42.1.47:38258] 404 GET /photo/:/transcode?height=552&url=http%3A%2F%2F127.0.0.1%3A443%2Flibrary%2Fmetadata%2F7232%2Fthumb%2F1637633625&width=368 (11 live) TLS GZIP 2ms 379 bytes (pipelined: 1)
Nov 23, 2021 10:36:01.718 [0x7f6613f75b38] Debug — It took 0.000000 ms to retrieve 142 items.
Nov 23, 2021 10:36:01.929 [0x7f6614684b38] Debug — Completed: [10.42.1.47:38252] 200 GET /hubs/metadata/7232/related?count=10&excludeElements=Actor%2CCollection%2CCountry%2CDirector%2CGenre%2CLabel%2CMood%2CPart%2CProducer%2CRole%2CSimilar%2CWriter%2CPhoto%2CVast%2CTopic&excludeFields=summary%2Ctagline%2Cfile&includeAugmentations=1&includeEmpty=1&includeExternalMetadata=1&includeLibraryPlaylists=1&includeMeta=1&includeRecentChannels=1&includeStations=1&includeTrailers=1&includeTypeFirst=1&libraryHubsOnly=1 (11 live) TLS GZIP 224ms 5775 bytes (pipelined: 3)

I recently moved my plex container from a stand alone docker vm, to a kubernetes cluster, which also happens to be using a traefik proxy. I wonder if there’s some correlation here?

I should add that I also have my Custom server access URLs in plex set for both the ip of the load balanced traefik ingress and the dns address that points to that ip.

Server Version#: Version 1.24.5.5173
Android Version#: 8.25.1.28703 (7bcafd43)

Hey Durden0,

I have a similar setup (without kubernetes), must be something related to traefik or how the referals are bein sent to the app.
It’s odd that it works perfectly on the web client.
Perhaps if other people who have it working with docker-traefik if they can share some of their config so we can see the difference.

Thanks if anyone could share some of their setup in docker with traefik.

Well, the guy who’s setup I copied when setting up my cluster reported he does not have this issue, and also runs traefik in front of plex. I’ve looked at it, and compared his setup to mine, but haven’t determined the source of the issue. Here’s his repo.

vs mine:

His setup is more complex than mine though.

I would also be interested to see if anyone more familiar with plex components could comment on what those warnings mean, why it’s looking at the 127.0.0.1 address and what is supposed to be happening with these requests.

I’ve been playing with the configuration in plex and noticed that it’s behavior changes based on the custom URL parameter. It’s very unclear what this parameter exactly pushes, I first thought this was just a way for the Cloud plex to redirect to your local plex server but I noticed that this parameter seems to be pushed to the local app.
My custom access url contained a list of possible ways to access the local plex server:
https://URL:443, https://URL:32400, https://IP:32400, https://IP:443
Changed it to:
https://IP:32400, https://URL:443
And it shows now the pictures however…
If i now access the local webpage it gives me the feedback it’s not connected directly to my server.

I’m totally lost with this parameter in plex, i don’t understand what it exactly does in plex and if it even checks each url in the list. If someone could enlighten me in the purpose and way how it works?

1 Like

hmm, i don’t use the 32400 port on either of my entries on the customer server access URLs,

https://192.168.1.1,https://my.url.com

Do you have traefik ingress setup to direct both port 32400 and 443 back to the plex endpoint? I only have it configured for 80/443 and 80 redirects to 443.

I’ll try playing around with this tomorrow and report back.

I only have ingress setup for port 443 on traefik, nothing regarding 32400.
But my docker compose file has 32400 as port off course and i’m running a virtual IP on my cluster (not that this really should matter).
Now there are 2 things i don’t really understand in the docker setup of plex.
In my docker-compose i had to add the parameter ADVERTISE_IP which is referring to my virtual IP and port 32400. So this might be related however what I don’t understand is what this parameter exactly does compared to configuring plex with custom access url.
In the beginning I had issues getting the plex access running in direct mode and had to fiddle with advertise_ip and custom access url’s to get it working. Hence it’s now the virtualip:32400.
What do you have configured in the docker file regarding this advertise ip?

First off, I just want to say, thank you. You helped solve my issue in a way that is acceptable to me.

So, silly me, forgot to check and see if my plex server was showing up as a “remote” server in the web browser before i started turning dials… but…

I tried trimming down my custom server access URLs to just

https://my.url.com

and like you, viola, thumbnails magically appear on mobile again. However, my server is also showing as ‘remote’ from the web browser. I have my Lan Networks set to contain my PC network, and two networks from within kubernetes (the pod network and the services network.

However, i tried confirming this behavior, changing my custom server access URL’s back to include the https://IP as well and can’t reproduce the thumbnail issue anymore. It seems to be caching it somewhere. I’ve even tried wiping the plex mobile app storage and re-logging in, still loads the thumbnails. I repeated this test on my wife’s phone as well. Same result, wouldn’t load thumbnails until i changed the setting. Soon as it changed, thumbnails show up and i can’t get them to not load anymore. So it must be caching them server side somewhere.

While, i don’t know why it’s showing up as Remote now, it’s an acceptable configuration for me for the moment, since I don’t care about bandwidth limits for WAN connections.

As for my ADVERTISE_IP, i’m not passing in that environment variable(only setting TZ), so I think it’s relying entirely on the Custom Server Access URL list for that. It also probably explains why, when i first deployed the container to Kubernetes, i had to proxy into the cluster to hit the pod directly instead using the ingress/svc address. Once i got in there, and set the custom server access url, it was available through the load balancer > traefik ingress point.

Let me know if I can be of further help in solving the Remote connection issue. For now i’m going to stand down on further troubleshooting of this on my own, but if you do find a solution, I’d be curious to hear about it. And of course if i stumble upon the answer later, i’ll update here to let you know.

1 Like

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.