[SOLVED] Cache says full when it's not.

[bprohs]

I'm totally confused...the more I research the problem the more confused I get!  I have 2 ssd cache drives.  one is 1TB the other is 250GB.  My cache device says I only have 625GB with 249 used and 0 free.  When I invoke the mover nothing seems to work.  My Sanbnzbd keeps pausing because it says my drive is full.  I'm hopelessly stuck...Thanks for any help!

[Squid]

Assuming that you've set up the cache pool as default, your 2 mismatched drives only will give you 250G of space (RAID1).

 

For why mover isn't working, you'd have to verify the use cache settings for the applicable shares

[bprohs]

So even though it says I have a cache pool, It will only write to the smaller one?  Should I disable the smaller one so it uses my 1TB drive?  

[Squid]

The default setting is RAID1, which means that every sector on each drive has to also be written to the other drive.  Since you've got a 1TB and a 250G the only way that's possible is that the net size is going to be 250G (btrfs does have a known problem with space reporting on mismatched drives, but the free space is always correct)

 

If you want the full 1TB, you can either convert to be a RAID0 (you'll get 1.25G) but lose the redundancy or get rid of the 250 and reset it all up (you'll get 1TB)

[bprohs]

When I try to convert it to RAID0, it just reverts back to Balance.

[John_M]

RAID0 will only give you 500 GB of space because, with striping, equal amounts from both devices are used. Single mode will give you 1.25 TB, but no redundancy.

 

"No balance found" is the normal, though somewhat confusing, status message.

 

[Squid]

Thanks.  My bad on the terminology

[amateur_wizard]

I'm having a similar issue so I figured I would post in the same thread.

 

Problem:

I have a 1TB cache, since I did some work on the server and some reformatting the it gets up to is 50% utilization. (It's a single disk)

It's formatted to XFS encrypted, same as the array. The logs does also read this as the cache being full.

 

I'm not sure what I did wrong.

Edited June 1, 2021 by amateur_wizard

[John_M]

1 hour ago, amateur_wizard said:

(It's a single disk)

 

It therefore has to be a different problem because you are not using btrfs RAID. Go to Tools -> Diagnostics and post the resulting zip file.

 

Edited June 1, 2021 by John_M
typo

[trurl]

2 hours ago, amateur_wizard said:

I'm having a similar issue so I figured I would post in the same thread.

I guess it's OK since the other user seems to have abandoned this thread. Instead of using the support thread of another user, in future it would be better to start your own thread with your diagnostics and a complete description of your problem (not really similar to the other users problem as noted).

 

[amateur_wizard]

Okay, I'll know for next time. I'm not up to speed on forum etiquette. 

 

Here are the diagnostics. It's been filled again to 50%.

arctvrvs-diagnostics-20210602-0544.zip

[John_M]

Check the minimum free space settings for each of your shares and your cache drive. They seem to be much too big - for example the share that's anonymised to M..a seems to be set to 500 GB and other shares to 1 TB, which gives the impression that the cache disk is permanently full. The best way to specify the large numbers required is to enter, say, "50GB" instead of "50000000000" because it's less error-prone.

[amateur_wizard]

That was silly, I didn't know that had a bearing on the cache. I totally forgot I set it that high or made a mistake with the number of 0's. 

 

Problem solved, thanks.