After getting a good preclear on the parity drive again.  I stopped the array, at this point everything was good.

I attempted to readd the parity drive and the cache drive.

 

As soon as I did this the web gui hung and when it came back it showed all array disks as missing and no parity drive available.

 

Unless you see anything in the syslog that looks like a smoking gun.  I'm thinking about attempting to use my motherboard's SATA ports in case there's a problem with my AOC-SAS2LP-MV8 (which would be strange considering it had been working for nearly a year, but anything's possible).

syslog-2014-01-20.zip

Jan 20 18:09:23 UNRAID emhttp: ckmbr: read: Input/output error

 

This is when the drives stopped responding.

What are your suggestions based on that error?

 

I've checked all cable connections.

I've actually tried two different SAS>SATA breakout cables on the MV8, both of which have locks on the SAS and SATA sides.

There's a new PSU and a new drive cage.  The PSU I think is fine as I'm not having any stability issues on anything else.  The cage has the 3 array drives in it.  The cache drive is on the old cage but still on the same controller.  And at least from what I can tell still shows up in the web gui while the others are missing.  That cache drive is also on the other SAS port on the MV8 controller than the array drives.

Test the system bare-metal.

So far so good, it's been running for a little over 24 hours now along with Plex installing.  I've yet to install any other plugins, I'll do that tonight and let it all burn in a little longer.

 

It's gone through a parity build and a parity check so far without error with the new SSD as the cache drive.

 

If it's all good by the weekend, I may try to reintroduce the virtualization piece with a fresh ESXi installation on a different flash drive.  Possibly just using ESXi 5.1 instead of 5.5 since I had 5.1 working for a year without problems until I added that new cache drive.

 

Thanks for your help so far.