Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Seagate 4TB “End-to-End” errors
#1
Hi, I’m reaching to the forum to get thoughts on this controversial topic. I have two Thecus N7510 NASs with 14 Seagate disks (Model # ST4000DM000) setup in a RAID 5 configuration and a daily rsync between the two NASs. I've had three disks that have been removed from the software RAID as failed with "end-to-end" errors within the last year, one disk in one chassis & two disks in the other chassis. Upon re-seating the drives, it passes both Short & long SMART tests and a bad block scan. One of the drives has experienced an "end-to-end" error event twice within the last month whereas one experienced a single “end-to-end” error last fall and nothing since.

What are “end-to-end” errors? From what I’ve determined from the internet this is a parity error between the data sent by the NAS and the harddrive’s (HD) cache (more below).

I think everyone would agree that if you get these over & over, it would indicate a HD issue. But what should we do if we only get one and it doesn’t happen again? And if it happens twice and the total “end-to-end” error is two but it doesn’t happen again, is that fine?

What would Seagate do in a warranty replacement? Would they just clear the SMART error information & send it back to me or would I get a new drive back?

Thecus says that they don’t have feedback from any other clients for this disk model. Seagate didn’t really say anything either way (warranty replace or not); they just said there was no firmware update for this drive (I already knew that).

My strategy so far is to monitor the disks & if mostly fine, I’ll replace all drives closer to the warranty expiry, and buy a couple new ones as well. The one that has experienced a failure twice, I’ll definitely warranty replace straight away on the third event.

I appreciate any opinions, thanks!!

=============

What is “end-to-end” errors & what do they mean? Here’s some opinions I’ve found on the internet:

“SMART attribute 184 (end-to-end error) indicates that the drive's on-PCB cache is going bad. The attribute is used to compare parity data between the original piece of data sent by the controller (in your PC/system) and the drive itself. This is not "bad system RAM", this is not "a bad motherboard", this is not "bad NAND flash on the HD". It's bad cache on the HD.”
<!-- m --><a class="postlink" href="http://www.tomshardware.com/answers/id-2234966/hard-disk-end-end-error.html">http://www.tomshardware.com/answers/id- ... error.html</a><!-- m -->

“First Smart values will never go away even if the issue doesn't currently exist.
Second all it saying is that data failed to be written and can be due to many different things such as but not including:
bad cables
windows failed to finish writing all data before powering off computer
you unplugged drive or computer before all data was written to disk
you reset the computer while it was writing to the drive
I can keep going on...
simple thing if the data is not changing then don't worry about it as 10 times is not much compared to the millions of write operations that have been done to the drive(each sector is a write op). There are bigger smart values to worry about like sector reallocation”
<!-- m --><a class="postlink" href="http://www.tomshardware.com/answers/id-2178847/smart-end-end-error-detection-data-drive-test-worry.html">http://www.tomshardware.com/answers/id- ... worry.html</a><!-- m -->

For completeness, this is where Seagate firmware updates & Seagate tools can be downloaded from:
<!-- m --><a class="postlink" href="https://apps1.seagate.com/downloads/request.html">https://apps1.seagate.com/downloads/request.html</a><!-- m -->
Reply
#2
Dear Sir,

According pervious experience, once HDD have any value from “end-to-end” which means this HDD have issue inside, and STRONG RECOMMEND to replace this HDD ASAP. Maybe you won't agree this point of view, but such issue happen on some user before.
Also Seagate & WD have Windows'utility to exam HDD, thus you could take HDD doing this test from Windows PC.
Yvon.
Reply
#3
Those Seagate ST4000DM000s are 'desktop' drives (even though Seagate touts them as good for NAS).
When you decide to replace them, you should consider NAS-class drives. NAS drives are designed to run 24/day, vibrate less, have better spindle bearings, and longer warranties.

Plus, when a desktop drive has a hard time reading a sector, it will try, try again for a longish period of time, longer than NAS drives will. If the RAID doesn't hear back from the HDD quickly enough, it will/might flag the HDD as bad. NAS-class HDDs report a fail-to-read-a-sector error sooner than desktop HDDs do - even though it might be able to successfully read it after more retries. The RAID will then rebuild that sector data, and is less likely to flag the drive as bad.
Thecus N5550 (CPU: D2550 @ 1.86GHz, BIOS: 2.05.14.5.cdv X64)
Disks: 5 x 3TB Seagate NAS-class; 2 x 3TB-RAID1 plus Hot Spare
Memory: 8125 MB (2 x 4 GB Corsair 204-Pin DDR3 SO-DIMM 1333)
Apps: Too many to count....
Reply
#4
Thanks! That's really good info. They might not have had the NAS drive out when I did this... but, I'll certainly replace them with the NAS drives. Cost differential is ~$15-20/drive. Enterprise drives are considerably more with a differential of about $150/drive. Note, Seagate did reply with more info and they said:

"End to End means that the data going into the cache didn't match what came out. This could be a drive related issue, and typically is such. However, considering the drives operated properly outside the NAS it's likely the drives are fine for their intended purpose.

We looked up the NAS and the hardware compatibility list, HCL, for the hard drives. While the drives listed are on the HCL, there is a footnote regarding the drives in the green section. "For business critical environments, we HIGHLY RECOMMEND using enterprise HDD'S."

Since the unit is a 7-bay NAS unit, We strongly recommends using enterprise-class drives; desktop-class drives may experience rotational vibration issues in boxes containing more than 4 drives. This means that the vibration caused by the spinning drives are compounded by being attached to the same drive cage, and may cause head height fluctuations. These fluctuations may cause the need to re-read the data multiple times, or in this case may cause the drive to incorrectly read the data, thus throwing an end-to-end error because even though the data is written correctly, when it re-reads it, the drive thinks that it differs from what was in cache."
Reply
#5
Good info, thanks for posting.
Reply


Possibly Related Threads...
Thread Author Replies Views Last Post
  Seagate 4TB “End-to-End” errors 0 1,521 Less than 1 minute ago
Last Post:

Forum Jump:


Users browsing this thread: 1 Guest(s)