Channel is Offline

fellowguy

Getting this in both jellyfin and in the ersatzTV Gui when I go to "Channels > Preview channels"

When I tail the log I get the following:

at ErsatzTV.Application.Playouts.BuildPlayoutHandler.Handle(BuildPlayout request, CancellationToken cancellationToken) in /source/ErsatzTV.Application/Playouts/Commands/BuildPlayoutHandler.cs:line 53
   at ErsatzTV.Services.WorkerService.ExecuteAsync(CancellationToken stoppingToken) in /source/ErsatzTV/Services/WorkerService.cs:line 64
2024-05-03 11:51:23.657 -04:00 [INF] Starting ts stream for channel 34
2024-05-03 11:51:23.667 -04:00 [INF] Starting HLS session for channel 34
2024-05-03 11:51:23.672 -04:00 [WRN] Error locating playout item {"$type":"UnableToLocatePlayoutItem"}. Will display error from "2024-05-03T11:51:23.6685875-04:00" to "2024-05-03T11:51:23.6685875-04:00"
2024-05-03 11:52:33.127 -04:00 [INF] Terminating HLS session for channel 34
2024-05-03 11:57:39.166 -04:00 [INF] Starting ts stream for channel 34
2024-05-03 11:57:39.176 -04:00 [INF] Starting HLS session for channel 34
2024-05-03 11:57:39.181 -04:00 [WRN] Error locating playout item {"$type":"UnableToLocatePlayoutItem"}. Will display error from "2024-05-03T11:57:39.1773793-04:00" to "2024-05-03T11:57:39.1773793-04:00"
2024-05-03 11:58:48.632 -04:00 [INF] Terminating HLS session for channel 34

I had this running on my Unraid server, but it was taking too many cycles, so I stood up a bare metal ubuntu server with docker and reinstalled, replicating the same process. So I'm not sure what I'm missing this go around.
I DO have a playout for the channel created set for any any any

I have installed jasongdove/ersatztv:latest-vaapi for Intel transcoding.

fellowguy

Log from GUI shows the following:

5/3/2024 12:10:07 PM -04:00 Information Terminating HLS session for channel 34 5/3/2024 12:08:57 PM -04:00 Warning Error locating playout item {"$type":"UnableToLocatePlayoutItem"}. Will display error from "2024-05-03T12:08:57.9217679-04:00" to "2024-05-03T12:08:57.9217679-04:00" 5/3/2024 12:08:57 PM -04:00 Information Starting HLS session for channel 34 5/3/2024 12:08:57 PM -04:00 Information Starting ts stream for channel 34 5/3/2024 12:05:38 PM -04:00 Information Terminating HLS session for channel 34 5/3/2024 12:04:36 PM -04:00 Warning Error locating playout item {"$type":"UnableToLocatePlayoutItem"}. Will display error from "2024-05-03T12:04:36.4904576-04:00" to "2024-05-03T12:04:36.4904576-04:00" 5/3/2024 12:04:36 PM -04:00 Information Starting HLS session for channel 34 5/3/2024 12:04:14 PM -04:00 Warning Failed to process background service request 5/3/2024 12:04:14 PM -04:00 Error An exception occurred while iterating over the results of a query for context type 'ErsatzTV.Infrastructure.Data.TvContext'. 5/3/2024 12:04:14 PM -04:00 Error Failed executing DbCommand (0ms) [Parameters=[@__buildPlayout_PlayoutId_0='?' (DbType = Int32)], CommandType='"Text"', CommandTimeout='30']

fellowguy

Can anyone help?

jason

Please paste the full error from the log file (etv config folder / logs subfolder) - it looks like there is an issue building the playout, which would cause the error you see at playback, but neither of the sections you pasted contain the full information about the error.

fellowguy

jason

https://pastebin.com/embed_js/gW4avLt8

Thank You!

jason

The logs indicate that you ran a develop release (which sometimes applies database changes), and are now back on an older and incompatible full release.

Switch your docker tag to jasongdove/ersatztv:develop-vaapi and stay there until the next full release (0.8.7), and things should work for you again.

fellowguy

jason

You're a bloody saint!

One last question.
Is this normal when watching ONE channel?

![https://ibb.co/C5x0mgs](https://)

CPU is pegged. Device is an i3-9100T CPU @ 3.10GHz with 16GB of ram. Linux installed to an NVME.

fellowguy

Well. It worked once. Closed out of the stream and went back into it, wanted to adjust the watermark as it was too big, and now I get this:

https://ibb.co/TRPVm9s

jason

fellowguy

I would confirm you have the /dev/dri device passed into the container, like --device /dev/dri:/dev/dri and also have VAAPI selected in the ETV FFmpeg Profile. That will enable hardware acceleration which should reduce CPU use.

If the ETV log doesn't show any errors when that happens, the 500 may be coming from Jellyfin. That /media/jellyfin/whatever URL in ETV just looks up the Jellyfin address and redirects the request (fffmpeg) there for the input file. The other way around this would be to include the exact same mounts in ETV's docker as Jellyfin's docker, assuming they are on the same machine. With that in place, ETV will stream from disk instead of through Jellyfin, though library updates and metadata will still come from Jellyfin.

fellowguy

jason

Like so?
https://ibb.co/YPHfXdh
Doesn't seem like its using the hardware. Even with these settings, it's using 95% CPU and occasional little stutter
Ended up selecting iHD, and it got better. Maxing at 80% CPU. Still a bit higher than I'd like. That's only with ONE channel streaming.

jason

fellowguy

I would check out the Troubleshooting page, VAAPI section. It tries all combinations of device and driver which may give you better performance than the default settings. You could also upload that info somewhere that I can take a look.