Hello

 

I recently just set up my unraid server. I had two brand new 4TB Black drives that I bought and I started to get Read Errors during a parity sync. I purchased a couple Seagate drives to replace them while I figured this out and have had no troubles since then with those disks.

 

I decided to run preclear on the blacks to see if there was any errors, there was not so I put them back in the array. Started writing to them again and again I get errors so I decided to RMA them. Got the new drives back from Western Digital earlier this week and started a preclear on both right away, ran for 37 hours and got no errors at all so I decided to put them into the array. Drives were blank, running a parity sync I start getting a whack of errors showing in the stats on the main screen on both drives. So I start thinking that maybe it's the SATA cable, the port on my motherboard, the power connector etc. that is causing this. Just to test I decided to put 2 older 1TB samsung drives into the array and start a new parity sync, that ran fine with no errors and those appear to be fine with no issues.

 

So my question is, does my system just not like something with the Black Drives, or is it possible my RMA'd drives are bad too, I find that hard to believe 4 drives would be the issue but with everything else running so smooth. I don't really know where to start or where to go from here. I am running a preclear again on one of the drives right now to see if there are any errors this time, and then I will put it back into the array again at that time, but thought I would see if any one on here had any ideas in the meantime. With other drives running perfect on the same cables and ports I find this so strange.

 

Thank you.

Post a syslog.

sorry, Here you go, I just stopped everything it was doing and ran one, only one of the Black drives is connected right now.

 

Thank you.

syslog.zip

Also ran a smart report on it and have attached it here.

 

Thanks Again.

smart_report.txt

I'm not familiar with the particular errors you are getting, sorry.  Hopefully an expert can chime in.

Smart report shows UDMA errors that could be due to bad cable connections.  I would check and reseat all cables as a first attempt.  Sorry not good at reading a syslog so didn't look there.

Blacks are 7200 and more power hungry. You didn't mentioned what PSU and how many disks you have in your server. An insufficient power supply may also be the problem.

Thanks for looking. I have the corsair tx-650 power supply http://www.newegg.ca/Product/Product.aspx?Item=N82E16817139020

 

I did buy it refurbished so not sure if that has anything to do with it, I know I probably shouldn't have done that since the PSU is one of the most important components. I didn't think it could be that cause all other disks seem to run fine off the same cable and power cable. I have the following running off of it

 

1 x 4TB WD Red Parity Drive

1 x 120gb Kingston SSD Cache Drive

2 x 4TB Seagate 4000DM000

1 x 1TB Samsung HD103

 

All of those are running good and then I have tried these two WD Black 4TB's, one in place of the 1TB samsung and the other in the last port on my motherboard.

 

I have checked all the cable connections and made sure that they are secure and have tried different SATA cables.

 

I let the parity sync run through last night and it finished with 1510 errors on that one disk, it said there was no parity sync errors. Running a parity check right now and it's about 40% through, there has been 9520 writes on this disk and 9510 errors, in the log I am getting those same read error, sector= as in the last log.

 

Just running an extended self SMART test right now to see what it shows. Maybe I should be trying another PSU though I don't know.

Did you 'neaten' up your computer by tying all the SATA cables together.  Doing so can cause cross-talk between the cables.  You should only tie the cables together if they are the shielded type.  (And they are not very common unless you really go looking specifically for them.) 

 

Edited to change 'cross' to 'cause'.

No my cables aren't tied together, they are touching in areas but it's hard to overcome that.

 

And it's not the power supply, I got a new 850watt today and it's still throwing up errors on those two drives. Thinking I will need to RMA them again and see, if it happens again with two more drives then I am at a loss, I will need to switch away from the blacks I guess.

 

Thanks for the suggestions so far, if anyone else thinks of anything else please let me know.

 

Cheers,