Seagate 8TB Shingled Drives in UnRAID


Recommended Posts

On 12/15/2018 at 5:56 PM, xerces8 said:

I have a 8TB Seagate Desktop drive (CystalDiskInfo says it is a ST8000DM004-2CX188, haven't shucked it yet).

The interesting thing is, CDM shows normal (as in "not SMR") values for 4K writing (same as for reading, about 1 MB/s), unlike about 10 MB/s, like for example my other 5TB SMR drive.

 

 

After filling it to about half and rerunning CDM, I got the typical values for 4K:  read about 0.5 MB/s, write 6 MB/s

Link to comment
  • 7 months later...

My 8TB Seagate Archive HDD (schucked 2020/02, Seagate Backup Plus Desktop Hub 4TB) has been sitting idle for almost three years. Lessons learned: you get what you pay for.

The disk has most likely not had more than 25TB written, and its already giving me all kinds of SMART errors.

Good news is that it contained nothing of value. Replacement disk won't be shingled that's for sure!

ST8000AS0002-1NA17Z

 

Check out values 5, 187, 198 & 198

Error 8341 occurred at disk power-on lifetime: 21731 hours (905 days + 11 hours)
  When the command that caused the error occurred, the device was active or idle.

  After command completion occurred, registers were:
  ER ST SC SN CL CH DH
  -- -- -- -- -- -- --
  40 53 00 ff ff ff 0f  Error: UNC at LBA = 0x0fffffff = 268435455

  Commands leading to the command that caused the error were:
  CR FR SC SN CL CH DH DC   Powered_Up_Time  Command/Feature_Name
  -- -- -- -- -- -- -- --  ----------------  --------------------
  60 00 08 ff ff ff 4f 00   6d+05:59:08.980  READ FPDMA QUEUED
  ef 10 02 00 00 00 a0 00   6d+05:59:08.980  SET FEATURES [Enable SATA feature]
  27 00 00 00 00 00 e0 00   6d+05:59:08.980  READ NATIVE MAX ADDRESS EXT [OBS-ACS-3]
  ec 00 00 00 00 00 a0 00   6d+05:59:08.979  IDENTIFY DEVICE
  ef 03 46 00 00 00 a0 00   6d+05:59:08.979  SET FEATURES [Set transfer mode]
#	ATTRIBUTE NAME	FLAG	VALUE	WORST	THRESHOLD	TYPE	UPDATED	FAILED	RAW VALUE
1	Raw read error rate	0x000f	115	099	006	Pre-fail	Always	Never	88515023
3	Spin up time	0x0003	091	090	000	Pre-fail	Always	Never	0
4	Start stop count	0x0032	100	100	020	Old age	Always	Never	597
5	Reallocated sector count	0x0033	100	100	010	Pre-fail	Always	Never	8
7	Seek error rate	0x000f	079	060	030	Pre-fail	Always	Never	97454053
9	Power on hours	0x0032	076	076	000	Old age	Always	Never	21732 (2y, 5m, 23d, 12h)
10	Spin retry count	0x0013	100	100	097	Pre-fail	Always	Never	0
12	Power cycle count	0x0032	100	100	020	Old age	Always	Never	51
183	Runtime bad block	0x0032	100	100	000	Old age	Always	Never	0
184	End-to-end error	0x0032	100	100	099	Old age	Always	Never	0
187	Reported uncorrect	0x0032	001	001	000	Old age	Always	Never	65535
188	Command timeout	0x0032	100	100	000	Old age	Always	Never	0
189	High fly writes	0x003a	100	100	000	Old age	Always	Never	0
190	Airflow temperature cel	0x0022	069	040	045	Old age	Always	In the past	31 (14 115 35 23 0)
191	G-sense error rate	0x0032	100	100	000	Old age	Always	Never	0
192	Power-off retract count	0x0032	100	100	000	Old age	Always	Never	805
193	Load cycle count	0x0032	098	098	000	Old age	Always	Never	4576
194	Temperature celsius	0x0022	031	060	000	Old age	Always	Never	31 (0 21 0 0 0)
195	Hardware ECC recovered	0x001a	115	099	000	Old age	Always	Never	88515023
197	Current pending sector	0x0012	100	090	000	Old age	Always	Never	256
198	Offline uncorrectable	0x0010	100	090	000	Old age	Offline	Never	256
199	UDMA CRC error count	0x003e	200	200	000	Old age	Always	Never	0
240	Head flying hours	0x0000	100	253	000	Old age	Offline	Never	2840 (65 240 0)
241	Total lbas written	0x0000	100	253	000	Old age	Offline	Never	18918662635
242	Total lbas read	0x0000	100	253	000	Old age	Offline	Never	6154007660

 

Edited by reboot81
  • Like 1
Link to comment
  • 4 months later...
On 1/23/2019 at 6:21 PM, wheel said:

So last month, I started having some issues writing to certain 8TB drives, but never took note of which specifically since no parity errors ever popped up before or after checks.  I was looking into adding more 8TBs to a 19-disk array this week when I realized I’ve been adding ST8000DM004-2CX188s left and right over the past couple of years.

 

SIX disks in a nineteen-disk array (not counting parity) are SMR drives (one is ST8000AS).

Ha ha. I've got 10 out of 30 of these. Been a solid 2-3 years so far. SMR is trash though. At the time, it was the best TB/$ out there, and I've been very lucky. I've since been buying 14TB Exos (ST14000NM0018-2H4101). These are doing really well! Eventually I'll be upgrading 1/3 of my server just to get rid of these SMR liabilities. Good luck!

Link to comment
9 minutes ago, falconexe said:

I've since been buying 14TB Exos (ST14000NM0018-2H4101). These are doing really well!

Well crap. According to this website, these ARE SMR. No where in the product documentation shows this too. Welp, I guess I'll go back to WD or Iron Wolf next time. 🤷‍♂️

 

https://blocksandfiles.com/2020/04/15/seagate-2-4-and-8tb-barracuda-and-desktop-hdd-smr/

 

https://www.seagate.com/www-content/datasheets/pdfs/exos-x-14-channel-DS1974-4-1812US-en_US.pdf

Link to comment
29 minutes ago, johnnie.black said:

Those 14TB Exos are CMR.

That's what I thought when I bought them, but these lines from that article are making me question that. I do believe you though...

 

"The Archive drives are for archiving and Exos drives are optimised for maximum storage capacity and the highest rack-space efficiency. Seagate documentation for the Exos and Archive HDDs explicitly spells out that they use SMR."

 

 

Edited by falconexe
Link to comment
Running a Benchmark on a SMR drive that's fresh from the wrapper using the DiskSpeed docker will show really high & flat speeds

That won't work with all SMR drives, SMR drives from Seagate and Toshiba look normal on speed tests, even before use, at least the ones I tested with, only WD drives do that, likely firmware related. 

 

 

 

Link to comment
5 hours ago, tucansam said:

You guys are freaking me out.

 

I have a whole truckload of 8TB Seagate Archives.

 

Is my server going to blow up randomly ?

Just stay on your S.M.A.R.T. reports and you'll be fine. First sign of trouble, swap it out immediately.

Edited by falconexe
Link to comment

Linear reads on SMR are the same as on non-SMR drives.

The biggest difference is in the random write test (10 times faster than non-SMR due to the track cache).

 

And hammering it with a _lot_ of read/write requests (similar results as non-SMR except for the occasional long delay).

 

Not exactly a one-click thing, but enough to get a few good hints.

Link to comment
11 hours ago, xerces8 said:

Linear reads on SMR are the same as on non-SMR drives.

Except WD SMR drives before they are written to, since the firmware knows there's nothing on those sectors in the platters they return the data from the cache, making them look much faster then they actually are.

Link to comment
On 8/20/2020 at 10:40 AM, falconexe said:

Just stay on your S.M.A.R.T. reports and you'll be fine. First sign of trouble, swap it out immediately.

 

Most of my drives have UDMA CRC errors, some have thousands, some have dozens, some have none.

 

A few have "Reported Uncorrected," value number 187, whatever that means.

 

Of 16 drives, those two errors are the only one that generate an orange line in the SMART value tables.

 

 

Link to comment
20 hours ago, tucansam said:

 

Most of my drives have UDMA CRC errors, some have thousands, some have dozens, some have none.

 

A few have "Reported Uncorrected," value number 187, whatever that means.

 

Of 16 drives, those two errors are the only one that generate an orange line in the SMART value tables.

 

 

Whenever I see anything light up in color, I immediately run a short and extended S.M.A.R.T. test. If results are verified, I then remove and replace that drive with a new pre-cleared drive.

 

These are the Usual Suspects in Order by Severity (My Opinion):

  • Current pending sector > 0

    • Reallocated sector count > 0

      • Offline uncorrectable > 0

If I see anything other than a fat ZERO here, I'm personally done with that drive. Others may say you can get away with it for a bit longer if you just have a low number of these and the metrics remain static, primarily reallocated sectors, but once there is 1, there will usually be more. Why take the risk? Especially when we are talking about SMR drives. I look for these opportunities to rid myself of these "mistake" drives. I got greedy a few years back with finding deals. For NAS type solutions, I would only get NAS type drives moving forward.

 

I'm sure others with vastly more experience with specific drive diagnostics will and should chime in here. I'm a veteran IT professional and I have learned a lot from this community when I joined in 2014. Everyone here is great, and they will always take care of you if you listen, have an open mind, and remain calm when crap hits the fan.

 

Hopefully this helps. If I were you, I would already have new drives shipping to my house and/or pre-clearing as we speak. And it goes without saying, I really hope you have good backups (onsite and cloud), and that you are already saving off this data. Good luck!

Edited by falconexe
  • Like 1
Link to comment

Thanks.  I will look into perhaps replacing them slowly, as funds become available.

 

If you can find me cloud storage for 88TB that doesn't cost and arm and a leg, I may consider it.  But I don't like the cloud, and I don't like having my data on someone else's server.  Hence me running unraid.

 

 

Link to comment
1 hour ago, tucansam said:

Thanks.  I will look into perhaps replacing them slowly, as funds become available.

 

If you can find me cloud storage for 88TB that doesn't cost and arm and a leg, I may consider it.  But I don't like the cloud, and I don't like having my data on someone else's server.  Hence me running unraid.

 

 

 

You should check out this thread on how Google Drives and Rclone can provide unlimited storage for around $12 a month. They don't actually require 5 users and they don't limit storage to 1TB when less than 5 users. Some users have well over 200 TBs.

 

 

https://gsuite.google.com/pricing.html

  • Thanks 1
Link to comment
17 minutes ago, tucansam said:

Google is the great Satan, we all know this.

 

Are you telling me storing data with them is safe?

 

I'm being serious.  Suggesting that I store data with google is very nerve-wracking to me.

 

 

Its fully encrypted where only you have the decryption keys.

Link to comment
17 minutes ago, tucansam said:

Are you telling me storing data with them is safe?

 

I'm being serious.  Suggesting that I store data with google is very nerve-wracking to me.

Safe is a relative term. The biggest issue would be if they suddenly decided to enforce their TOS, in which case you would either abandon them, or pay the higher price. Like BRIT said, you can set it up so only you can see the content of the data, assuming you don't have the attention of a three letter agency.

 

If you use them as the third tier of a sound backup strategy, you shouldn't have any issues, other than the massive amount of bandwidth needed to get the backup started.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.