UDMA CRC errors on new UnRaid server - Advise

[beachbum]

Hello Unraid Community,

 

I'm very excited to join this community and as such I'd like to ask for some advice.

 

I've built an UnRaid server of 3 16 TB Seagate "Amazon Refurbished" drives, 2 1TB SSD for cache and 1 256GB Nvme drive. Following instructions from Space Invader in youtube i have run a Pre-clear on my 3 16TB Drives.

 

However at the beginning of the process one of my drives gave out a SMART count of 4 CRC errors, while the other 2 no errors at all.

In the end the pre-clear process was completed with "Preclear finished Successfully" message.

I panicked and i requested a replacement drive from the seller. I have read some horror stories about these "Amazon drives" so i thought i had gotten a bad drive.

While they process the RMA for a replacement drive I have been doing some research and found that CRC points to a "faulty" Sata cable and no so much a faulty drive.

My question is, should i just replace the SATA cable and hope to not see the errors again or should i proceed with the replacement drive ?

It's a rather a painstaking process to replace the drive due to its location on the server so i rather not replace it if it's not faulty.

 

While i understand that ultimately the decision relies on me, i'm just trying to open a forum to discuss the gravity of a CRC error so that maybe me, and others can learn more about it.

 

I welcome all comments on this issue and thank you in advance for your time...

 

Glad to be a part of this community.

 

[apandey]

CRC errors can be due to anything no being stable in your data pipe. Basically data being sent from computer doesn't check out when written to drive. It can be controller, cables (SATA/power), drive itself. More often than not, its the cables / connections as they are the most fragile part in the chain. 

 

I would start by checking / reseating / swapping / replacing cables. That is most likely all you need

 

If you want to be doubly sure, you can run an extended smart test on the drive. This test runs completely within the drive firmware and is thus immune to any external factors. If that comes out well, your drive is OK. It's an expensive test to run, so I will still just start with cables

[beachbum]

11 hours ago, apandey said:

 

I would start by checking / reseating / swapping / replacing cables. That is most likely all you need

 

If you want to be doubly sure, you can run an extended smart test on the drive. This test runs completely within the drive firmware and is thus immune to any external factors. If that comes out well, your drive is OK. It's an expensive test to run, so I will still just start with cables

thank you for your response.

as you suggested I replaced the sata cable and started a new pre clear process on just that one drive.

 

I ran a standard smart test that retuned no errors.

 

question, when you say the “extended smart test” it’s expensive… what do you mean by that ?

is it posible to cause damage to the drive or just takes too long ?

 

As I read what I wrote I’m thinking maybe you meant “extensive” ?

 

thanks again.

[itimpi]

51 minutes ago, beachbum said:

when you say the “extended smart test” it’s expensive… what do you mean by that ?

is it posible to cause damage to the drive or just takes too long ?

Just that it takes a long time to run.  Typically something like 1-2 hours per TB of the disk size.