Setup is a HP microserver running unRAID 5.0 beta13

 

Short story is that I've been using this perfectly fine for months, no errors reported. I check my Main page today and find the attached failure.png image showing the parity disk with a blinking red light and number of errors. A quick scan through the syslog revealed problems trying to spin up the disk and a number of read/write errors (I have attached this in a zip file as it was too large)

 

So I tried to run a SMART test on the drive but it would not detect it. I opened the case and checked the SATA and power cables to make sure it wasn't loosened somehow and stopped the array. After it attempted to spin the drive up again, the drive is now detected as blue - a new parity disk (see after.png) - and I can run SMART (report output attached), but I really have no idea what to do next and am scared to restart the array and have it rebuild the parity if the drive is failing.

 

Can anyone offer some advice on next steps and what they think is going on from the attached syslog and SMART reports?

smart.txt

syslog-jan-23rd-2011-22.56pm.zip

Here also attached is the extra data from the syslog in the past hour following stopping the array.

 

Thank you to anyone that can help!

syslog-jan-23rd-2011-23.42pm.txt

The SMART report looks fine. Rebuild parity and then run a parity check. Then post a new SMART report

Thanks again for your help. I restarted the array and let it rebuild the parity data (it only took all day ) and it seems to be running without error or incident.

 

I've attached the updated SMART report.

 

Any ideas what might have caused this strange glitch powering up and not being able to read/write to the drive properly?

smart.txt

I would:

Double check your power supply is large enough and your sata cables are of good quality and making a good connection.

 

If you have  a spare drive I would swap out the parity and run the preclear script on it a couple of times to make sure its ok.

 

 

 

Thanks, I have the standard PSU that came with the microserver, I would imagine it must be up to the task given that it's a four swap-bay system, I think a fair few people on here have systems built around them?

 

I will double check the connections again though and keep an eye on things; I wish I could afford a spare 3Tb drive but not with prices as they are!

ya, perhaps not the best time to buy a 3TB drive

I thought when I purchased a couple spare 2TB drives for $79 several months ago I was being too frivolous with my money - now I'm very glad.

 

Your other option would be to shutdown the array and run preclear on the drive, then bring it back online and rebuild parity if its ok (but that means likely a couple days without the array and then a day without parity protection while it rebuilds).

Thanks, I think I will not risk that for now. If it happens again I definitely will consider it.

 

I got both 3TB drives in my array at the same time from the same shop for £100 just before prices skyrocketed (cheapest are around £140 here now), so its more concerning that if they are from the same batch they could both fail at around the same time. It was a lucky bargain though!

The SMART report looks OK. Make sure you check SMART for failures. Run a parity check after parity is rebuilt.

Parity check has also completed without incident or error, thanks for your help guys. I'll keep a regular eye on it.