I skipped a song and everything disappeared. I have tried reloading and no changes
whatever seems fixed now.
It seems that, regardless of user actions, the AutoDJ and Playlist tabs fail to fetch the channel playlist from the API server, sometimes with high probability, but at random.
When they fail, the dev console usually contains this traceback:
vendor-36d7c2a1.js:26 TypeError: Cannot read property 'message' of null at app-e5cfb458.js:formatted:1065 at vendor-36d7c2a1.js:26 at l (vendor-36d7c2a1.js:26) at vendor-36d7c2a1.js:26 at d.$eval (vendor-36d7c2a1.js:26) at d.$digest (vendor-36d7c2a1.js:26) at d.$apply (vendor-36d7c2a1.js:26) at o (vendor-36d7c2a1.js:26) at m (vendor-36d7c2a1.js:26) at XMLHttpRequest.x (vendor-36d7c2a1.js:26)
Sometimes they do succeed and actually load the playlist, but even if that works, once the AutoDJ switches to the next song, it very often fails to load the new player.
Could this be looked into?
Thanks for bringing this to our attention. I will look into this and get back to you all. I would guess there is some woes with the ElasticSearch cluster.
Wondering if you are making any progress on this problem ? It’s a real pain.
Performed some maintenance on the cluster tonight, and will be monitoring today to see what changes result.
WELL - Tonight the music didn’t stop at all for me in a 7 hour stream
Fingers crossed it keeps up
It all fell apart again tonight hun. Music stopped a number of times and wouldn’t play the next song again. Had to skip forwards about 8 songs before it started playing again.
Annoying I can’t reply 3 times in a row !! Posting what I WAS going to post on this comment instead.
I was wondering if any further work was going to happen on this problem. It still happens, albeit less often than before it seems. Is there any way I can run a log of some sort to send you or anything that will help ?
Last night it stopped playing at least 10 times in 6 hours, and I had to repeatedly skip songs (at least 2 or 3 each time) until it started playing again. Then it would only play 1 or 2 tracks and then stop again. It’s really getting to be a PITA !!!
It’s unlikely this will get fixed permanently anytime soon. It looks like there are internet woes at the datacenter where playlists live during the times you reported these issues. Since the datacenter connection is over the internet, any sort of latency or networking issues will cause issues with playlist loads. When the internet is stable these issues do not occur.
Ouch, well that sucks. Still, kudos on the work you’ve put in. It’s always nice when developers deals with problems instead of ignoring them, like some companies tends to do sometimes. (I won’t name names.)
Suggestion - Is it not possible to create a LOCAL file on my own machine and for that to be used to access the playlist ? I realise it’s probably not right this moment. I’m just saying that if Nightbot could read from a file we each have on our own computer, it could be used as a backup backup !!! i.e. look on the database first and if no access to that THEN it picks a track from a local file. Just a thought. It could be stored in the same folder that the current_song.txt file is stored in.
So basically i have a problem: When i request a Youtube Music Video it won‘t be played in the stream like i do the requests etc but it won‘t start and when i go back to my autodj settings it starts the video but the song is still not playing in my stream
You have to have the browser page open for it to play. The volume is controlled by the “desktop” option on OBS and also the volume vontrol on the page.
this has also been a long running problem for me. nightbot seems to temporarily lose the play list and gives a database error. after closing the app for 15+ minutes fixes it. but having to do that multiple times a stream is very annoying. I have been googling for work around periodically for the last month
is nightbot just not well suited to playing music?
These issues really only affect the playlist, not song requests themselves. We have yet to fully address the playlist woes, but we hope to address it in the near future. We’re still slowly working on infrastructure changes behind the scenes which are needed to improve this condition.
This is still an on going problem. Is there any news for this? Also for me it happens between the hours of 5pm est and 9pm est. Works fine throughout the day an night. Is this a server related issue due to many people using the auto DJ function at this time?
We have been slowly working towards infrastructure upgrades behind the scenes, and hopefully will have a more permanent fix for this in the near future.