currently I do see that the cache is locking out because it was below its minimal thresh hold

Check you shares floor setting or free up some space on the pool, appdata for example is set to 200GB, and the pool currently has less then 100GB free, so no writes will be possible to that share, also note that there are files from appadata on disks 4 and 5, they should be moved to the pool.

7 hours ago, JorgeB said:

appadata on disks 4 and 5

 

I will look at this again , I thought I got them all. 

 

Yes , after I posted diag I noticed I was under the threshold and fixed that. 

 

Is the way I am putting cache in the drive path vs saying /user. Is that correct ? 

I also have the sab completed / incomplete on cache , I am wondering if I should move that. 

15 hours ago, ninjaneer68 said:

Is the way I am putting cache in the drive path vs saying /user. Is that correct ?

Since the share is now set to cache=only you can use either.

I spent time working on this over the weekend I will post setting tomorrow evening after work and get some opinionss

This post is a follow-up to the previous issue. The original post in this thread, was corrupt cache drive that needed reformatted but as we went through the journey, there was other things that needed fixed.
During the resolution process, I realized that I hadn't set up some settings optimally for my use cases, as highlight by help from others in this post and discord.

Issue: Docker wouldn't start, cache free space was below threshold ( put cache in read-only mode )

Why: my cache drive was full, all downloads went to the cache drive and I did not have array as secondary storage set up

Resolution: I did a series of changes over time based off of learning and feedback from the mods in this forum. The list of changes is listed below

 

• Reformatted the corrupt cache drive.
• Updated the share where the Sab downloads go to the following settings ( see figure 2 )
  • Primary storage: cache
  • secondary storage: arrary
  • mover action: cache >> array
    • I set the above the way I did so the new downloads will hit the cache drive and the drive speed will not be the bottleneck for downloading, I don't need the files to live there long term though, hence the mover action to the array.
• The share that has appdata, I did the same as the above, except for mover action. I want the app data to live on the cache, but if it accidently fills up I don't want to see all the containers crash ( see figure 3 )
  • same as above except mover action is different
  • mover action: arary >> cache 
• One of the other mistake adding to the issue, for the sabnzb container, I hard coded the path for the cache vs user.  ( see figure 1 )
  • what I had was /mnt/cache  vs having /mnt/user 
  • at the time, I had it this way, because I did not have the shares configured correctly to handle this for me. when the cache drive went into read only mode, everything started having issues. 
     
• Figure 1 - 
• Figure 2 - 
• Figure 3

Edited December 27, 20232 yr by ninjaneer68