[Unraid OS version 6.8.0-rc7 available]

Quote

 

md/unraid driver
Introduced "multi-stream" support:

Reads on devices which are not being written should run at full speed.  In addition, if you have set the md_write_method tunable to "reconstruct write", then while writing, if any read streams are detected, the write method is switched to "read/modifywrite".

The "stripe pool" resource is automatically shared evenly between all active streams.

 

Can confirm this feature works. however, is there possibly a method/setting to revert behaviour to work as before?

i have situations where i want the mover to complete quicker rather than delay due to a single active stream (that uses Kb's odf data verus about 50MB's of transfer speed that is unavailable for the mover.

[[6.8.0-RC5] Cache drive full - BTRFS mismatch with GUI]

Confirmed issues is resolved. thus configuration issue only.

 

@johnnie.black Thank you sir for picking it up!

 

Might not be a bad idea to update the GUI to show entered numbers in more readable units. (i was caught off guard that it is multiplying the entered value by 1024...

 

 

[[6.8.0-RC5] Cache drive full - BTRFS mismatch with GUI]

Changed Status to Retest

[[6.8.0-RC5] Cache drive full - BTRFS mismatch with GUI]

30 minutes ago, johnnie.black said:

This is the allocated space, not the available space, btrfs allocates new chunks as needed.

 

 

 

very interesting (although why only allocate 130GB when 147GB is used according to the GUI and DF -H?

30 minutes ago, johnnie.black said:

 

 

That's because you're cache floor is set to 20TB, remove 3 zeros.

 

DOH

i didn't read the guide properly. changed it to 20GB now. will test normal operation and report back

 

thank you!