Hi,

I have 2 arrays: Tower (5b11) and Tower2 (6rc3).

I updated Tower to 6rc3. Before this Tower was consistently copying data to my new array (Tower2) at 30-35MBs. After the uptdate it only achieves 6MBs.

Pls, can anyone help me out of this.

Both arrays replicate critical data (as backup) and I need to achieve asap those 35MBs.

Also, after the update parity check gets stuck at 7%.

None of both things happens to me in my other array.

Rgds.

EDIT: When parity check starts it runs at up to 100MBs speed, but when it arrives at 7/8% this is where it ends up:

	
Elapsed time: 	31 minutes 	
Current position: 	160 GB (8.0 %) 	
Estimated speed: 	3.6 MB/sec

What can cause this?

I run SMART selft-test and all the disks passed it.

 

syslog.txt

I'm sorry, but this syslog did not include the parity check!  It's last recorded event was at May 29 19:39:33, when you clicked the Log button.  All 3 drives were spun down well before that.  To help, we would need to see a syslog with parity check, captured after it slows down.

Hi, I attach my syslog and a screen capture.

<a href="http://s236.photobucket.com/user/razzta01/media/unraid%20-%20varios/errors/array%20tower1_zps9pkulzyr.png.html" </a>

Let me know what you see.

Rgds.

syslog_2.txt

The parity check runs fine, for the first 26 minutes.  THEN... every single parity block is wrong and has to be corrected!  I have to assume that previously, it was doing the same, correcting parity for each block, until you stopped it, and this time, it moved quickly over the corrected region until it reached the uncorrected.  It's picking up where it left off.

 

The implication is that this is not the original parity disk for those 2 data drives.  Is it possibly the wrong drive?  Or was there another data drive with it, that was removed when you created this new server?  That would make parity completely invalid.  At this point, you should stop and rebuild parity.  There are no drive errors, so the hardware looks fine.

 

If these parity corrections were occurring during the file transfers, then that would explain why the transfers were so slow.

Hi, thx fir quick support! I am rebuilding data now and I al getting normal speed numbers again.

Rgds.