Well I thought I would get a little more input on this but maybe I'm the only one that plays with storage spaces. Have a backup drive on the way to rebuild my storage. Still going to use storage spaces but provisioning will be set to fixed this time. The one thing I wondered is why a storage spaces is defaulted to thin. Whatever drives I would want to use for mass storage I'm trying to use whatever space I can get. I understand setting thin provisioning for vms because you can set many to one drive. So anyways new setup. Mirrored refs fixed provisioning. Hopefully good to go.
Storage spaces low capacity while drives are half full
Ok I've researched and seen this but no real answer other than start over. Have a mirrored refs storage pool on win 10 pro. Pool was created with 2 6tb drives and then 2 8tb drives were added. This all before any data was actually written to them. That was last year. No failures or hiccups until now. Server is accessed remotely and not always looked after but getting a low capacity warning to add 2 drives and shows all 4 drives are 99.99 percent full. I over provisioned the pool originally to 24tb for future expansion. With 14tb of actual storage it's not even using 7tb yet. Optimization will not run for the gui. Running in powershell it kicks out that the volume is too full. System restore past last major update did nothing and system restore was checked to make sure that the mirrored drive was not listed causing mirrored snapshots. ![74581-img-20210303-053648.jpg]![74591-img-20210303-053355.jpg] : /api/attachments/74581-img-20210303-053648.jpg?platform=QnA : /api/attachments/74591-img-20210303-053355.jpg?platform=QnA
I meet a similar case before, however, it seems there's limited useful information yet, I will keep researching on it, if there's any update, I'll feedback as soon as possible.
I believe I have found the root of the problem but no answer to it. Windows 10 workstation pro was used originally to create the storage pool. The gui does not give an option for thin/thick provisioning. Default is thin. What's happening is there are alot of blocks not written back to zero due to the thin provisioning and leading to the crazy numbers it's spitting out. I have not found a way to recover any of these blocks.
I know the retrim command should fix this. However there is not enough empty space for it to run. I'm going to add 2 drives to the pool tonight and see if I can optimize and trim then remove the drives from the pool. If this works I'll set a task up with powershell to run the optimize command on a schedule. Unless anyone else has input on this.
Added 2 vhds to it remotely using an attached 2tb external drive. Did not solve any issues. File system is under a constant back up which is good. So going to try to enable dedupe to see if that can give the space to run retrim. If it doesn't I'm going to have to rebuild the array from a backup.
Sign in to comment
Sort by: Most helpful
Thanks for your sharing!
Dunno if this will help but similar problem solved
Has this issue recurred since you set it up with fixed provisioning? Running into the same issue and I might have to offload all my data and re-create the storage space again. I assume you used PowerShell to set it to fixed?
Sign in to comment