Hi -

 

I'm running version: 5.0-rc16c. Parity + 5 data disks + a cache drive. There is very little customization on here - I have set it up, and just let it run. It's rock solid.

 

I pulled up my www interface today just to see what % capacity I am at, and noticed some errors showing on both my parity drive and one of the data drives. I am attaching a SMART report for both, as well as a Syslog.

 

(The only concerning that I see is a Current_Pending_Sector count of 11 on my parity drive, in addition to a bunch of "Tower kernel: md: disk4 read error, sector=115085512 (Errors)" errors reported for both drives.)

 

I'm running a parity check now (had been > 120 days) and we'll see what it looks like when it's completed.

 

I don't mind replacing one or both of these drives if I need to....but I'm wondering if I really need to yet.

 

Any advice on things to watch for, or other advice would be appreciated.

 

thanks

Matt

syslog.zip

Data_Disk4.txt

Parity_Disk0.txt

Pending sectors indicate an unreadable or barely readable sector that has been marked for relocation, but has not be reallocated yet.

 

Both of your drives have them.

 

I would try a parity check or two and see if it gets worse or better. unRAID is able to reconstruct an unreadable sector and write the data back to the drive, which should trigger the reallocation.

 

I have seen "false" reallocated sectors which completely clear themselves with no apparent ill-effects. I think it is a firmware bug.

 

If you can get them to reallocate and the numbers hold steady for 3 consecutive parity checks, I would continue to monitor it but feel the drives are safe for now. But if every parity check results in new or worsening conditions, I'd return them.

 

Might want to look at the HGST drives. They are more reliable based on studies I've seen.

 

 

One other thing -- upgrade to the stable v5.0.5 release.  The reason there were release "candidates" was to identify bugs that could be killed before the final release.    And even after v5.0 final, there was a .01, .02, .03, .04, and then a .05 release (the current v5 "stable").

 

I'm attaching the SMART reports for the two drives after running through 2 parity checks. Those parity checks were finished with no sync errors reported.

 

I haven't upgraded to the 5.05 version, but will try that today if i have a few minutes.

 

thanks again for the input.

Parity_Disk0_new.txt

Data_Disk4_new.txt

I haven't upgraded to the 5.05 version, but will try that today if i have a few minutes.

 

Just to be clear, that's not going to really "do" anything to resolve your issue;  it's just a good idea to be using the stable version so you're not trying to resolve problems that have already been fixed 

I'm attaching the SMART reports for the two drives after running through 2 parity checks. Those parity checks were finished with no sync errors reported.

 

I haven't upgraded to the 5.05 version, but will try that today if i have a few minutes.

 

thanks again for the input.

 

These smart reports both look fine. No reallocated sectors and no pending sectors. As I mentioned, I've seen pending sectors clear with no apparent ill effects.

 

I would not take them out of service, but would monitor them. Maybe do weekly parity checks for 3-4 weeks and make sure all continues to look good.