[glow=red,2,300][/glow]Hi all,

so I have been upgrading one of my unRAID servers and added several new drives. I have been adding the drives one at a time, pre-clear, add the drive to the array, pre-clear next drive, add to the array, etc. While finishing the last pre-clear the previous drive added gave me a red ball and is now disabled. There is really nothing written to the drive and my suspicion is either a bad cable or the PSU (Corsair TX650M) . My question is, once the last drive is done pre-clearing, I am thinking I will replace the red ball drive with that drive, then once the rebuild is done I will telnet into the server and perform the trust array commands below and see if I get any more red balls on that drive.

The array is backed up onto another unRAID server so losing a small amount of data on this server will just mean copying it back from the other server.

 

Before doing this I am going to also swap the drives in the cages they are in so that I can eliminate any cable issues and if needs be, replace the sata cables.

 

I have attached the system log below.

Thoughts?

syslog-2014-12-14.txt

As an update, I took the safe course of action and placed the newly cleared drive into the array and I am letting the array rebuild. I did swap the two drives so that I can check to see if the cabling is the issue. Once the array has been rebuilt I intend to assign the drive that red balled back into the array and hopefully have no issues after that. I really believe the drive is fine but only time will tell.

As an update, the drive I swapped into the slot that the original red balled drive was in has now given me a red ball too. This confirms my original suspicion that the cable or hot swap cage may be the issue.

The drive I swapped out is currently finishing a pre-clear which so far has been successful, I will swap it back into the slot that the latest drive failed in, once I replace the sata cable. Hopefully this will fix the issue.

Well as the final update on this post. I replaced the sata cable and even moved the drive from an add-in sata controller card to the onboard card and I was still getting massive errors. I finally bit the bullet and bought a HX850 to replace the one year old CM 650 PSU and I finally have a stable system which ran pre-clear on the new drive and then a full parity check after the new drive was added with no errors.

So it seems 8 drives for the CM650 pushed it over the edge, even though the PSU thread on these forums indicate that PSU should handle many more drives than 8.