[SOLVED] SMART 184 End-to-End_Error

IsaacKuo · 10-15-2018, 03:45 PM

I'm getting the following error on a 5TB hard drive I just bought off of eBay. I was aware when I bought it that it wasn't brand new or anything.

Code:

The following warning/error was logged by the smartd daemon:

Device: /dev/sdc [SAT], Failed SMART usage Attribute: 184 End-to-End_Error.

Device info:
ST5000DM000-1FK178, S/N:W4J1CD5K, WWN:5-000c50-09b44c774, FW:CC49, 5.00 TB

For details see host's SYSLOG.

You can also use the smartctl utility for further investigation.
The original message about this issue was sent at Thu Oct 11 03:08:53 2018 CDT
Another message will be sent in 24 hours if the problem persists.

From what I've read, this error indicates there may be an imminent failure at any time. I'm okay with that - I have my data backed up on other hard drives (although they may have issues also; see below).

But what's more troubling to me is the possibility that the data is being corrupted on the way in or out without me knowing it. As I understand it, the end-to-end error means that some sort of on board cache checksums don't check out. I've read it suggested to turn off the write cache, or something like that.

Should I disable the write cache (and/or other caches)? How do I do that? I tried to search around to figure out how to do that, and it just confused me as to what to actually do.

Currently, I am using this 5TB as my main nfs share on my main file server, with the other drives being used to backup data from it via rsync. If disabling a write cache can make it more trustworthy, then I'd like to do that. I'd appreciate any help figuring out how to do that.

Thanks!

(My other big hard drives also have issues:

My 3TB drive has - 24 Currently unreadable (pending) sectors

My 2TB drive has - 1 Currently unreadable (pending) sectors

I'm not sure what to do about that either.)

All of my big drives have just one big ext4 partition.

smallpond · 10-15-2018, 04:55 PM

I doubt you need to disable the drive cache. The drive has to stage data in its RAM while reading or writing the platters anyway so if the RAM is bad, disabling cache will not prevent corruption. More likely, you have either media errors or a bad read or write channel. Replace the disk. If you do want to disable cache, the command would be:

Code:

hdparm -W0 /dev/sda

syg00 · 10-15-2018, 05:43 PM

That's a SMR disk - the were really flakey when first released due to the vendor specific nature of the fiddle they use to increase capacity. According to this (first search hit) those have been pulled out of NAS boxes and resold.

It's possible the error is spurious, but how would you know ...

IsaacKuo · 10-15-2018, 06:01 PM

Cool, thanks for the info! The command I used was:

Code:

hdparm -W 0 /dev/sdc

The eBay listing did explicitly describe the drive as being taken from an NAS after a short period of use. I'm okay with that. I pulled my 3TB drive from an enclosure (I had been given it as a gift, so it didn't cost me any money). So basically, I'm aware that I can't really trust these drives as much as a typical new retail internal hard drive.

Fundamentally, I have backups of my files on the other drives, but they're scattered among a whole bunch of smaller drives rather than a few big ones.