Forgive me as I am new to the unRAID but not to servers/computers;

 

I've added a newly purchased 3TB Western Digital Green drive to my array on an HP MicroServer N40L as my parity drive (Currently have 1x 3TB Parity, 2x 1TB Data).  Before hand I did preclear the drive successfully with no errors.  Added it as my parity and ran the Check (originally had the defaulted checkbox of fix parity errors) and it resulted in " Last checked on Mon Nov 10 00:51:43 2012 GMT, finding 0 errors. "

I thought all was splending until I looked at the Error colum on my parity drive and it showed 12928.  This made me feel uneasy.

I read through the forums and looks like we recommend running another Check on parity; I did so but this time unchecking the fix parity errors box.

This resulted in "Last checked on Mon Nov 12 00:51:43 2012 GMT, finding 0 errors."

Fantastic I thought!  But I again go take a glance at the Error column on parity to find the number 6781.

Is this an area of concern?  I want to make sure my parity is perfect before I go about adding more drives to my array ...

 

Here is the system log I've been able to attach.

 

Any thoughts, comments, ideas, would be extremely helpful.

Lately all I've been saying to friends and colleagues is "UnRAID FTW"

syslog.txt

Forgot to mention, This is the latest unRAID v5.0rc8 as I needed 3TB support.

 

I am zipping the entire syslog file if that helps.  I've also added the smartlog for the 3tb parity drive.

 

 

syslog.zip

smartlog.txt

Forgot to mention, This is the latest unRAID v5.0rc8 as I needed 3TB support.

 

I am zipping the entire syslog file if that helps.  I've also added the smartlog for the 3tb parity drive.

According to the smart report, one sector has been re-allocated and none are pending re-allocation.

 

You should be fine.  Just keep an eye on it for the next few months.  (the "errors" you see in the web-management-console are "read" errors.)  But it appears as if the one un-readable sector was re-allocated already.

 

Joe L.

Looking further (in the syslog) the errors are in seeking to a sector, not an un-readable sector.

Nov 11 15:28:24 Tower kernel: hdb: dma_intr: status=0x51 { DriveReady SeekComplete Error }

Nov 11 15:28:24 Tower kernel: hdb: dma_intr: error=0x40 { UncorrectableError }, LBAsect=8574736, sector=8574336

 

Unless the power supply is marginal for the number of disks you have attached to its 12Volt rail, or there is mechanical vibration causing seeks to fail, you might just have a disk with issues.

 

Here is one thread (from a google search) with a similar error.  His resolution, a new power supply:

http://www.linuxquestions.org/questions/linux-hardware-18/hdd-errors-hda-dma_intr-status%3D0x51-%7B-driveready-seekcomplete-error-%7D-480669/

 

Joe L.

Thanks for the tips Joe L , I will certainly be checking this out.

Because I have the parity drive currently in a 3rd party hotswap bay, I'm going to remove it from that and connect it directly in hopes it could just be the hotswap bay hardware.

If that fails to resolve the errors, I will continue the path to try and exchange the power supply.  That is going to be a total pain if it resorts to that, and might be more cost effective to replace the entire MicroServer.

 

Your comments are much appreciated.

Well I tried removing the 3rd party hotswap bay that was used on my parity drive and rerunning parity check but I still get thousands of those same read errors

Trying a new power supply is going to be difficult as all the cables on the MicroServer N40L are neatly cable managed..and I have not been able to find a replacement on the cheap.

 

Do you think this is critical? Will my unraid /data still be able to rebuild in loss of an actual drive?

I've been able to copy terrabytes of information without an issue.

 

5 green drives running on the HP N40L seem to be a common thing users run.. I don't think that should be an issue on the 150w power supply

I happened to purchase a second 3TB disk and I will begin preclearing that disk shortly.