Hi,

 

I have been moving a lot of files around on my server just using my Windows 10 VM. After a while the VM hung and I when I checked the unraid console the 'Fix Common Problems' plugin said there was an issue with my cache.

 

Samsung_SSD_850_EVO_mSATA_500GB_S33HNX0H612268V

 

The cache drive was completely full so I am guessing this has something to do with it. None of my VMs will boot now of course leaving me in a predicament given this is a workstation.

 

cache (Samsung_SSD_850_EVO_mSATA_500GB_S33HNX0H612268V) has file system errors (No file system (no btrfs UUID))

 

Any ideas on how to fix this problem?

 

Thanks.

 

 

Adam

server-diagnostics-20161026-2216.zip

Try your luck with the BTFRS recovery tools:

 

https://lime-technology.com/wiki/index.php/Check_Disk_Filesystems#Drives_formatted_with_BTRFS

Thank you.

 

My array is all reiser so should I be using the reiserfsck?

 

I tried this as per the link 'reiserfsck --check /dev/md1' but it says the /dev/md1 is 'Not a Directory'.

 

Should I be entering something here specific to my array?

 

Thank you.

 

Adam

You array is reiserfs but your cache is btrfs, at least that's what it's set to.

 

 

 

 

Thanks I was just checking the disk log and saw it mention that.

 

So I don't recall selecting a FS when I added the cache disk, is BTRFS the default? I obviously don't want to nuke the disk by running the wrong check

BTFRS is the default for cache.

 

If you're not sure you can click on cache, change the cache filesystem to auto and start array to confirm, but the setting wouldn't change by itself.

Thanks you, not sure what I would do without your help

Ok, I set to auto and then restarted as suggested.

 

FS is marked as reiser!

 

Fix common problems now shows this error:

 

All writes to your cache enabled shares are being redirected to your array. If this is a transient situation, you can ignore this, otherwise adjust your cache floor settings here:Global Share Settings or adjust the frequency of the mover running:Scheduler Settings or purchase a larger cache drive

 

I have another SSD I can add as a second cache drive. Will that help?

 

Adam

Reiserfs only supports one device, you need to change to BTFRS to support a pool (cache need to be formatted for that)

 

I'm guessing that the plugin is telling you your cache disk is full.

yes the cache disk is full.

 

I think what happened is I saw it was full and added a second SSD that wasn't formatted correctly and it threw the cache. So anyway I am now back to only having a full cache.

 

How do I free up space or get it to finish updating the array or basically back to normal?

 

Thanks,

 

Adam

Run the mover.

I think what happened is I saw it was full and added a second SSD that wasn't formatted correctly and it threw the cache.

 

BTW, this is why the cache filesystem changed to BTRFS, setting more than one slot for cache changes the filesystem to BTFRS, even if you then change back to 1 slot it remains as BTFRS.

Ok that makes sense.

 

So ideally I need to get things back to normal, back up my VMs and then reformat as BTRFS (and then restore the VMS) if I want to add another cache disk?

 

Will adding another cache disk improve speed or is it purely for increasing the available size?

 

Thanks again.

 

Adam

So ideally I need to get things back to normal, back up my VMs and then reformat as BTRFS (and then restore the VMS) if I want to add another cache disk?

 

Yes, only BTFRS supports a pool.

 

Will adding another cache disk improve speed or is it purely for increasing the available size?

 

By default a RAID1 mirror is created for redundancy, no added space or speed, but it can be converted to other profiles, see here:

http://lime-technology.com/forum/index.php?topic=48508.msg484482#msg484482

 

Thanks so much for your help. I now have an error free array and my VMs are working.

 

However the unraid GUI is REALLY slow (e.g. the Apps pages takes minutes to load and for the update spinner to go away, searching for Krusader returns no results...) and so is my Win10 VM. I am guessing something is not quite right. Any tips on where to start troubleshooting?

server-diagnostics-20161027-1615.zip

You Win10 VM seems to be using a lot of CPU, if you shut it down does the GUI start behaving normally?

Unfortunately it makes no difference. Both VMs are idle and the array is also not being used at the moment.

I'm guessing that the plugin is telling you your cache disk is full.

I had to look up that warning, as it didn't really make sense.  It however makes perfect sense when the other part of it (the actual short warning) is included in the quoted text:

 

Cache Disk free space is less than the cache floor setting

 

All writes to your cache enabled shares are being redirected to your array.  If this is a transient situation, you can ignore this, otherwise adjust your cache floor settings here: .... or adjust the frequency of the mover running....

Yes, my bad 

 

Any ideas why my array is running like a dog? I can't even install Krusader to back up my VMs and try a clean install.

 

It is very frustrating, everything was running super speedy until I ran into this problem...

Ok, I figured it out (I think). I installed a new router and forgot to change the gateway port so unRAID couldn't accest github. Doh! 

 

Hopefully this has cured the other slowdowns too, we will see.