Closed ridgekuhn closed 1 year ago
Thanks! That's looking good. 15 minutes is probably good.
Also, it's missing one piece. Restarting the "radico8-youtube" service won't fix the youtube stream unfortunately. You have to hit some API endpoint first that tells youtube you're starting a stream, then restarting the "radico8-youtube" service will fix the youtube stream.
Also, I could add a CHANNEL_NAME to that env file. The env file currently only has 1 line:
PASS=<youtube-stream-token>
I'm working on documentation rn...
Ok, I updated the main README with how you setup a server. It includes the content of those env files along with what you have in this PR (though it isn't merged yet).
I guess I'll stop procrastinating and try testing out that livestreaming api: https://developers.google.com/youtube/v3/live/getting-started
I'm merging this now so I can build on it.
My bad, that makes sense. Were you were doing that manually in the browser?
Yeah, I was clicking through the youtube studio to do that when the stream went down.
I actually just figured out an easier fix that doesn't require a timer. Here's the API parameter description: https://developers.google.com/youtube/v3/live/docs/liveBroadcasts#contentDetails.enableAutoStop
enableAutoStop defaults to true and you can't configure it through the UI, which kinda sucks. But you can set it to false through the API. False means the broadcast will never go down unless you manually stop it. If the server is having issues, the stream might show buffering for a while until radico8 is in a good state again.
I've been testing it for the past 30min-hour and the broadcast isn't going down. So that's great. I'll add some scripting/documentation in the repo then close that issue one issue.
Thanks for looking into it though. It gave me motivation to investigate too.
I'll probably end up removing the youtube healthcheck too. Systemd has been doing a good job at restarting radico8/radico8-youtube when it is in a bad state. And there isn't a need for it now since I found that API setting.
Oh that’s perfect. Weird that u can’t set that through the web UI. Now that you’re using the API, I wonder if it might be better to do just do the health check that way, but I read last night that doing so can use up a lot of API calls that count towards your account limit so that’s why I went w the web scraping route. Anyway, happy to help out even if it was just for motivation, lol!
On Feb 25, 2023, at 7:43 PM, Alan Morgan @.***> wrote:
Thanks for looking into it though. It gave me motivation to investigate too.
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.
oh funny timing, u beat me to it about the API health check, lol
On Feb 25, 2023, at 7:50 PM, Alan Morgan @.***> wrote:
I'll probably end up removing the youtube healthcheck too. Systemd has been doing a good job at restarting radico8/radico8-youtube when it is in a bad state. And there isn't a need for it now since I found that API setting.
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.
FYI, I haven't tested this at all but I saw this was on your todo list in the YT chat and thought I might be able to help. I think it should work, but I wasn't sure about the contents of the
youtube.env
file. Could u add documentation for that or an example .env to the repo? I went w a variable namedCHANNEL_NAME
to check against the URL.