I have three 2T drives in my unraid configuration.  The machine is slowing down in its response.  I go into a telnet session and try and ls /mnt/disk1 and it takes quite a while, the same happens on /mnt/disk2. I assume one or more drives are going bad.  What is the process to figure out which drive and replace it?

M-

Drives generally don't degrade that way. It seems more likely to be a RAM issue.

Are the drives spinning up?

The problem was the parity drive was failing reads.  This seems to be a shortcoming of unraid since the whole array slowed down and there was no yellow or red indicator on the status screen.  I replaced the parity drive and rebuilt it (this took 6 hours for 2T) and things are back up and running.  I would suggest that a drive failing reads should be treated almost the same as one failing writes.

 

M-

Write errors are indicated on the Main page under Device Status in the Error column.

Yes I understand that write errors are tracked in unraid.  Its just that this array slowed way down over time and there was no error indication on the web page.  When I did a smartctl it showed that 900 sectors had been moved. 

M-

Write errors are indicated on the Main page under Device Status in the Error column.

 

I meant read errors. A write error will cause the disk to become disabled. The disk must indicate a read error has occurred for unRAID to log it.

Write errors are indicated on the Main page under Device Status in the Error column.

 

I meant read errors. A write error will cause the disk to become disabled. The disk must indicate a read error has occurred for unRAID to log it.

and read errors are show, even for the parity disk.

I think a better approach would be that if reads become slow or unraid detects that lots of sectors are being redirected, that this is indicated on the webpage.  BTW don't get me wrong, I think unraid rocks, but this experience showed how unraid could be improved.