UDMA CRC error count on replacement drive

[rdlind]

I had a drive that was triggering a failed array because of UDMA CRC error count issues.  My first attempt was to reseat all cables and rebuild the array with the disk that had the UDMA CRC error count issue.  Seemed fine for awhile (the disk is at 75% so its mainly read from) but then failed the array again so I purchased a new disk.  My setup includes 3x4 disk hot swap containers sitting in a server case that feed up to a card in a Dell R610 (I can get exact product numbers if needed).  Anyway I was only running 8 disks, filling up 2 of the 3 hot swap bays (so total of 8 rotational 3.5" drives), from what I've read UDMA CRC error count tend to be bad connection type issues, so to be safe I put this new drive in one of the bays I have not used and rebuilt the array.  It went perfectly.  Couple days post rebuild I open up the Unraid dashboard to find that this new drive in a new slot has triggered 1 UDMA CRC error count.  What is going on? Is there anyway the data itself on that drive could have some issue causing this?  I am really confused.

 

Thank you for any help you can provide.

[trurl]

Go to Tools - Diagnostics and attach the complete Diagnostics ZIP file to your NEXT post in this thread.

[rdlind]

As requested.  Replacement drive now showing UDMA error is ST8000DM004.

 

Thanks again for any help.

akashic-diagnostics-20200924-2024.zip

[Frank1940]

Make sure that you have not bundled your SATA data cables with cable ties to make the inside of the case 'more pretty'.  Bundling the cables can lead to cross-talk problems which are another cause of CRC errors.

[trurl]

Likely nothing wrong with the original disk. And nothing apparently wrong with this one.

 

An occasional small number of CRC errors, just go to the Dashboard, click on the warning, and acknowledge it. It will warn again if it increases. If it continues to increase, especially if it is not a small number, then try to fix the connection problem.

 

No need to replace the disk.

 

 

[rdlind]

Thank you both for your help.

 

Frank1940, I am running SFF-8088 to SATA cables that run to a LSI 9201-16e host card in my R610, I may take a look and see if there is some very strange situation of a power cable or something sitting too close to the cables, but by design these cables are crammed together.

 

trurl, that is definitely what my research led me to believe also.  My original disk (WD Red 8TB) ran for quite awhile (maybe a couple years) before the CRC errors started.  Sadly they continued to grow larger and larger until unraid determined that the drive was no longer valid.  I will continue to watch my replacement drive and hope this was a terribly timed fluke.  No other drive (3x8TB, 4x4TB) has even registered one CRC error in their years of use.  This mystery of why the disk that contains "Disk 2"'s data,  no matter where they are connected, are the only drives I have ever had have CRC errors will continue to elude me.

 

Thank you both again for your time and information.