Hey guys, I have been getting RANDOM disk errors for awhile now. Each time i've gotten these errors, I have pulled the specific drive and gotten a brand new one. Each time it's a different drive, different bay, etc. I will attach my log here and you guys can tell me what you think. I REALLY don't think that i'm constantly getting a bad hard drive...The first 2 times this happened, it was my Parity Drive. Now it's one of my Data Drives.....

 

I have a SUPERMICRO CSE-743TQ-865B-SQ Server case. I'm starting to wonder if it's the HotSwap panel? I'm currently using a SAS Card too (Model# is: LSi 9211-8i in IT mode)

 

PS. The drive that is NOW having disk errors ends in "WG8025TP". I also want to just reiterate that I have been having this issue constantly with now THREE DIFFERENT DRIVES...So I do not believe that this is a hard drive issue anymore...

 

Thank you for the help in advance.

 

 

UPDATE!

 

I shut unRAID down, swapped the "errored disk" into a different bay..then restarted the array...Now i'm able to see the Smart report for that drive and everything looks fine...So what is actually happening here?

 

I'll attach the new smart report for that drive as well. (AFTER REBOOT)

 

This is what i'm looking at, is there a way to get rid of this X without PreClearing that drive?

 

 

syslog.2.txt

syslog.txt

syslog.1.txt

ST8000DM004-2CX188_WG800K3E-20170917-0833.txt

ST8000DM004-2CX188_WG8025TP-20170917-0835.txt

ST8000DM004-2CX188_WG801405-20170917-0833.txt

WDC_WD80EFAX-68LHPN0_7SGEEU7C-20170917-0833.txt

ST8000DM004-2CX188_WCT01187-20170917-0833.txt

ST8000DM004-2CX188_WG8025TP-20170917-0902 (AFTER REBOOT).txt

Edited September 17, 20178 yr by Josh Scheretie

47 minutes ago, Josh Scheretie said:

This is what i'm looking at, is there a way to get rid of this X without PreClearing that drive?

 

You don't need to preclear, just rebuild.

 

Disks look fine, so the multiple errors would point to other hardware problem, cables, controller, power supply, etc.

@johnnie.black Okay, I will take a look and see if I can narrow this down. I appreciate the reply!

6 minutes ago, johnnie.black said:

 

You don't need to preclear, just rebuild.

 

Disks look fine, so the multiple errors would point to other hardware problem, cables, controller, power supply, etc.

What are the exact steps to Rebuild the Array if using all the same drives...

Stop the array, unassign disk1, start array, stop array, re-assign disk, start one more time to begin the rebuild.

14 minutes ago, Josh Scheretie said:

@johnnie.black Okay, I will take a look and see if I can narrow this down. I appreciate the reply!

 

Disk cabling issues are the #1 cause of unRAID disks getting errors and dropping offline. And even once the cabling is perfect, when you go to swap out a disk, it is incredibly easy to inadvertently touch a cable on another drive enough to create an intermittent connection. Hours, days, or even months later that bad connection will cause problems.

 

I highly recommend hot-swap style cages, like the CSE-M35T-1B. You wire these cages once, and once burned in, swapping disks in and out is easy. And you never come in contact with the disk cables again (at least until the next major upgrade).

12 minutes ago, bjp999 said:

 

Disk cabling issues are the #1 cause of unRAID disks getting errors and dropping offline. And even once the cabling is perfect, when you go to swap out a disk, it is incredibly easy to inadvertently touch a cable on another drive enough to create an intermittent connection. Hours, days, or even months later that bad connection will cause problems.

 

I highly recommend hot-swap style cages, like the CSE-M35T-1B. You wire these cages once, and once burned in, swapping disks in and out is easy. And you never come in contact with the disk cables again (at least until the next major upgrade).

 

Whole heartedly agree with every word, one other tip is label all the cables at both ends, makes trouble shooting way easier, and locking SATA cables are a must imho.

35 minutes ago, CHBMB said:

 

Whole heartedly agree with every word, one other tip is label all the cables at both ends, makes trouble shooting way easier, and locking SATA cables are a must imho.

 

Locking is great if they are supported. But I've had locking cables give problems, so not an absolute guarantee. If not locking, at least good friction is needed. Older cables tend to loose that friction over time and slide on/off too easily, and even if not technically broken, they are not good to use.

 

The best connection is one that has been burned-in with heavy use and not disturbed by human hands afterwards.

 

I used to label before going to drive cages and SAS HBSs. It's certainly easier to manage sets of 4 than single cables. What I'll often do is dedicate all 4 ports of a SAS connector to one 5in3, leaving one spot empty, which I will fill with a motherboard ports.

 

For example (A-D are from SAS cables, M is from motherboard).

 

AAAAM

BBBBM

 

or, for a more elaborate setup

 

MMMMD

AAAAD

BBBBD

CCCCD

 

This is actually an excellent setup. Typically leaves 2 motherboard ports for SSDs (which I don't mount in the hot-swap cages).

 

 

Makes it easier to stay organized.

I only use SAS cables, and all of mine are labelled both ends.  I love my label maker....

Thank you so much for the points guys! I'm planning on taking out my Hotswap Backplane and wiring them direct to see if it's possibly the backplane panel