cache mover "no space left"

[nunu22]

Hello everybody.

I'am testing unRAID 6.11.5.

I have created an unprotected array device with 2 disks of 8T and an other of 4T.

i have 2 mains docker application: resilio and emby.

All was working fine ... but as all was working too fine, i have decided to test the cache feature with a SSD disk of 500gb 😁

 

The SSD start to fill up, and once it was full ... tranferts were stopped (from resilio side: no left space), and the cache flush fails (mover)
So i have checked on forum etc... but without success.

 

So i join the diagnostic, if somebody can help me ?!

 

 

 

Edited April 4, 2023 by nunu22

[itimpi]

You need to set the Minimum Free Space settings to be larger than the biggest file you expect to cache (we normally suggest 2x this value to give some headway).    This tells Unraid when to stop using the pool for caching or when to switch to another drive on the arrsy.

 

this setting exists in the settings both for any pool, and also for each share.

 

This is covered here and here in the online documentation accessible via the ‘Manual’ link at the bottom of the GUI or the DOCS link at the top of each forum page.

[JorgeB]

You need to set a minimum free space for the share(s), should be about twice the largest file size you expect to copy there.

[nunu22]

Here space occupation of the array :

 

 

 

Edited April 3, 2023 by nunu22

[JorgeB]

Click on the share or shares you are using and set "minimum free space" as mentioned.

[nunu22]

i have just change it since first answer :

 

[trurl]

You will have to run mover manually to get files moved from cache, or wait for it to run at the scheduled time. 

 

Best if you don't attempt to cache initial data load, or any file transfer larger than cache.

 

It is impossible to move from faster cache to slower array as fast as you can write to cache, and trying to move while still writing to cache only makes it worse. Mover is intended for idle time. 

[nunu22]

some thing strange ....
i have put 40G for min free space (appdata and media shares).

I have deleted files directly in the cache disk to free space ....

all my dockers are stopped .... now i see my cache to be filled from disk1 to cache...

 

 

You're right regarding the speed "issue" between SSD cache and HDD....

i have missed something hughe ?!

 

 

[trurl]

Your user share anonymized as m---a is set cache:prefer. Prefer means prefer to keep all files for the share on cache. These will not be moved to the array. Set this share to cache:yes.

 

Typically, the only shares you want cache:prefer are appdata, domains, system. These shares currently have files on the array, so after you make room on cache by moving m---a to the array, you will have to disable Docker and VM Manager in Settings so these shares can be moved to cache where they belong.

 

After you get all that done post new diagnostics, you might still have some cleanup to do with those shares

[trurl]

You can see how much of each disk is used by each user share by clicking Compute... for the share on the User Shares page. Ultimately you want all of appdata, domains, system shares on cache, and all of other shares on the array. Then you can cache that m---a share if you want, but it must be set cache:yes or it will not get moved to the array.

[nunu22]

in the GUI documentation:

 

i haven't see the tiny différences between the 2 sentences (maybe a bold can help)....

 

thanks a lot ... cache is flushing ...

[itimpi]

It looks like the wording of this setting and its associated values is going to be completely redone for the 6.12 release.    The underlying functionality will be the same but hopefully the reworked presentation will be much clearer and less likely to be misunderstood by new users.

[nunu22]

fine

 

i have read several times theses sentences ...

BUT in gui we have necessary informations ...

 

 

 

thnaks a lot for your fast support !!!