I seem to have an issue since moving to 5.0 where mover is choosing one drive to fill instead of spreading the load. I can share dirves and write directly to the array, but mover has always worked correctly in the past.

 

Any thoughts? Look at disk "5"; share level is at "2" as it has always been... Got some big drives ready and waiting.

 

Thanks in advance.

 

Are your shares all set to include all drives?

 

Is the split level the same on all of the shares?

 

Post a picture of the Shares page with the details for each share.

 

All are the same - attachments will follow. Thanks

Thanks for looking

LASTLY...

So if you write directly to /mnt/user/TV it goes to a different drive but letting the mover do it everything goes to disk 5? Just asking for clarification...

 

Can you show a "ls" of /mnt/cache ?

DOS spoken here... Is this what you requested?

Ok, so your cache drive is currently empty, was hoping there was something there that might tell why everything is going to disk 5.

 

Did you answer my first question?

So if you write directly to /mnt/user/TV it goes to a different drive but letting the mover do it everything goes to disk 5? Just asking for clarification...

Sorry, no, I did not.

 

I can share (samba) any drive and write directly to it - no issue.

 

In fact, I removed ~ 300GB to a local drive then wrote it to disk 12. As you see, disk 5 is again the preferred target.

 

Thanks!

Albert

Ok, so the issue seems it is with unraid deciding what disk to use and not so much the mover script... Sadly I haven't explored that far into unraid so someone else will have to continue

No worries, thanks for your attention!

 

I suppose I could set a threshold for the drive and see if that takes it out of the loop. Thought I'd ask the community before hosing something.

Instead of setting a threshold, exclude it from the share and test that way.

will do, thanks

No Joy - moved about 80GB of data to the "Movies" share from a W7 workstation last night. The data was on the cache drive as expected. When I woke this morning unRaid had ignored the exclusion of disk 5, filled it up again, then moved on to disk 12 which is in fact the largest empty space (see original post for screen shot comparison)

 

Is the new version "seeing" disk 5 as larger than it is and ignoring exclusions?

 

WTF

 

Thanks!

Are you saying that there is no Movies share already present on the disk being used in error?  Just checking as if there is the relevant folder already there your split level might be coming into play and forcing the new file/folders to be on that drive.

The share (folders) exist on all drives (2000+ titles). Something in 5.0 appears to be ignoring the remaining array to publish to disk 5. After filling it again, the next largest, disk 12 was used correctly.

 

I even excluded disk 5 from all shares before attempting to store more data last night - also ignored.

 

I have three unraid servers and several years of maintenance without issue. No issues with rc16 were noted.

 

Thanks

What is the full path of the files being written?