Emby docker disappeared in Unraid

[dcruff 15]

I have been setting up Emby on Unraid to replace my Windows server.

I just finished moving over all of my media and massaging the media files to get the metadata and artwork correct and was getting ready to make it go live.  Then I had a problem with my 1 TB Samsung SSD cache drive which now needs to be reformatted or replaced.  When the cache went down and refused to mount, I lost my Emby docker as well as my Krusader (File Manager) docker - Gone.  Now, I'm wondering how much data was being stored in the SSD cache drive.  I believe that I have most of my meta data and images stored inside my media folders.  

Can I just reinstall the Emby docker and see everything come back up?

[Luke 42078]

3 minutes ago, dcruff said:

 

Can I just reinstall the Emby docker and see everything come back up?

Hi, at this point yes I would try that. Let us know how you get on. Thanks.

[CassTG 113]

1 hour ago, dcruff said:

I have been setting up Emby on Unraid to replace my Windows server.

I just finished moving over all of my media and massaging the media files to get the metadata and artwork correct and was getting ready to make it go live.  Then I had a problem with my 1 TB Samsung SSD cache drive which now needs to be reformatted or replaced.  When the cache went down and refused to mount, I lost my Emby docker as well as my Krusader (File Manager) docker - Gone.  Now, I'm wondering how much data was being stored in the SSD cache drive.  I believe that I have most of my meta data and images stored inside my media folders.  

Can I just reinstall the Emby docker and see everything come back up?

Depends how you set up the cache drive and how you set up the folders you created. Each folder has a different setting for how you want cache to be handled, if you had bulked transferred everything across with cache sitting in front of yall our main drives/folders it may not of moved everything across as there is a setting for Move operations (i.e move on shutdown, spinup length of time etc), been a while since i used unraid but it might not have necessarily moved everything across depending on the time between transfer and failure.

If i recall appdata folder is  where it stores dockers is normally cached , which is what you want if your drives are all spinners as obviously faster access equals better emby performance DB wise. 

Not that it helps you now but there is a great plugin in the community store for backup, it will create backups of all dockers and vms via cron and place them elsewhere (spinners non cached for example) so once you are reset up you can set that to run once a day then if the same happens then least you can just redeploy the previous backup image pretty quickly

[dcruff 15]

Thanks.  I decided to bulk load from my external backups and I shut down the cache because it was causing a lot of headaches with the upload of 14 TB of media files.  The 'mover' was too limited and was wearing out.  Actually, I didn't know that a typical SSD is limited in it's lifetime's write ability and it was heating up and getting errors.  Apparently, I shouldn't have shut down the cache for all of the shares, because that's when I lost my dockers and now my 1 TB Samsung SSD is Unmoutable.  I've just reinstalled the Emby docker and set up my libraries again.  I was hoping that Emby would find my previous configurations, but it did not and now I am setting things up again.  I'm hoping that it has recognized the nfo files and pictures stored in the media folders.  I'll know shortly.

I'll see if I can get the cache to work again once I am all done because I really do want any performance enhancements that it can offer while Emby is streaming.

[CassTG 113]

Yes that would be the best outcome, i never ran emby on my unraid as i host in the cloud, but thats a lot of IOPS to be hammering, good thing with the cache in unraid is as long as you shutdown the array correctly you can force moves and add / remove the cache drive, so yeah i would definitely transfer everything across and let emby finish its scan before adding back the cache drive. I had a 960Gb Corsair mp510 nvme drive and warnings would show up circa 50-60 degrees and that was with a heatsink on.

 

Yes as mentioned the Mover which handles the transfer of files to and from the cache drive is not always invoked until the specific setting you set in the unraid config, so if the array was not stopped and the data "moved" out then in theroy its sitting on the knackered drive.

 

 Unraid is a great system and i enjoyed it whilst i used it but  after digging deeper in to it i i felt it's docker handling was a bit convulted and bolted on and i had odd issues with windows vm performance, switched to proxmox to run my  vm's with one vm setup for the Docker stack and found that this solution suited my needs better

[dcruff 15]

I have reinstalled my Emby Docker and had to set up my Media Categories again, but it was fairly painless.  It seems to be running very smoothly right now.  I also reformatted the cache drive but have not enabled its use on any of the shares yet.  I'm still concerned about the cache though.  It's not being accessed at all but its temperature is 93 degrees.  I know it can get hotter, but that still seems warm when not operating.

 

[CassTG 113]

Is that 93C or 93F, if the later then that is about right for operational temp, if the former i would chuck it.

[TestRuff 0]

9 hours ago, CassTG said:

Is that 93C or 93F, if the later then that is about right for operational temp, if the former i would chuck it.

93F which I learned is not really that hot.  It appears to be working well right now.  Apparently these SSD's run kind of warm.

 

[CassTG 113]

1 hour ago, TestRuff said:

93F which I learned is not really that hot.  It appears to be working well right now.  Apparently these SSD's run kind of warm.

 

Yeah 93F is fine, mine when being hammered would sit at late 30's mkd 40's celcius, only thing is when they get hot for long the lifespan reduces unlike dram which does not mind a bit of heat