SOLVED Is smartctl accurate for SAS drives? - [update] It is

[rwillett]

I've just finished building my NAS server, thanks for the help in getting the H310 card reflashed.

I've just run smartctl on my old but new SAS Seagate ES.2 Constellation disks and they appear to have 31,000 hours logged on them.

Code:

root@freenas[/tmp]# smartctl -a /dev/da0
smartctl 6.6 2017-11-05 r4594 [FreeBSD 11.2-STABLE amd64] (local build)
Copyright (C) 2002-17, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF INFORMATION SECTION ===
Vendor:               SEAGATE
Product:              ST330006CLAR3000
Revision:             YS18
User Capacity:        3,000,592,981,360 bytes [3.00 TB]
Logical block size:   520 bytes
Rotation Rate:        7200 rpm
Form Factor:          3.5 inches
Logical Unit id:      0x5000c50055e890e3
Serial number:        Z295T4A4    00009306KPYX
Device type:          disk
Transport protocol:   SAS (SPL-3)
Local Time is:        Thu Apr 18 11:48:31 2019 PDT
SMART support is:     Available - device has SMART capability.
SMART support is:     Enabled
Temperature Warning:  Disabled or Not Supported

=== START OF READ SMART DATA SECTION ===
SMART Health Status: OK

Current Drive Temperature:     43 C
Drive Trip Temperature:        68 C

Manufactured in week 52 of year 2012
Specified cycle count over device lifetime:  10000
Accumulated start-stop cycles:  131
Specified load-unload count over device lifetime:  300000
Accumulated load-unload cycles:  131
Elements in grown defect list: 0

Vendor (Seagate) cache information
  Blocks sent to initiator = 4148158124
  Blocks received from initiator = 2892790048
  Blocks read from cache and sent to initiator = 2517390898
  Number of read and write commands whose size <= segment size = 1873146723
  Number of read and write commands whose size > segment size = 0

Vendor (Seagate/Hitachi) factory information
  number of hours powered up = 31762.37
  number of minutes until next internal SMART test = 27

Error counter log:
           Errors Corrected by           Total   Correction     Gigabytes    Total
               ECC          rereads/    errors   algorithm      processed    uncorrected
           fast | delayed   rewrites  corrected  invocations   [10^9 bytes]  errors
read:   3437540563        0         0  3437540563          0     474749.026           0
write:         0        0         0         0          0     113503.555           0
verify: 110169080        0         0  110169080          0    2712207.870           0

Non-medium error count:       18

SMART Self-test log
Num  Test              Status                 segment  LifeTime  LBA_first_err [SK ASC ASQ]
     Description                              number   (hours)
# 1  Background short  Completed                   -   31760                 - [-   -    -]
# 2  Background short  Completed                   -   31760                 - [-   -    -]
# 3  Background short  Completed                   -   31758                 - [-   -    -]
# 4  Background short  Completed                   -   31757                 - [-   -    -]

Long (extended) Self Test duration: 6 seconds [0.1 minutes]

root@freenas[/tmp]#

I suspect that smartctl is accurate, but does anyone know if this could be wrong or is it likely that by 'new' disks are actually pulled from a data centre or NAS?

It's been a difficult day all day :(

Rob

 

[jgreco]

So you want to know if your 2012-model drive that says it was built in 2012 and has 3.62 years of lifetime logged with 131 start-stop cycles is... what?

I'd call it a lightly used data drive of some sort that spent an unusual amount of time turned off.

 

[Mlovelace]

Hmm... Logical block size: 520 bytes this smells like a NetApp drive or perhaps the slightly more elusive EMC drive. You'll probably want to convert it to 512 bytes if you're intending on using it with freeNAS.

 

[rwillett]

The disks were sold to me as "new and unused". I knew they were old but I wasn't bothered about the age per se, rather that they were new. If they've been lost in a cupboard and now found thats fine by me.

My question was whether smartctl is accurate for all disks, Your responses indicate that you think so.

I didn't pick up on the Logical Block Size, 520 bytes is odd, I think I misread it as 512.

I don't know if 31,000 hours is a lot these days, but the usage was supposed to be zero hours. Seagate indicate that these disks have a MTBF of 1.2M hours, so 31,000 hours doesn't appear to bethat bad.

Rob

 

[Mlovelace]

The disks were sold to me as "new and unused". I knew they were old but I wasn't bothered about the age per se, rather that they were new. If they've been lost in a cupboard and now found thats fine by me.

My question was whether smartctl is accurate for all disks, Your responses indicate that you think so.

I didn't pick up on the Logical Block Size, 520 bytes is odd, I think I misread it as 512.

I don't know if 31,000 hours is a lot these days, but the usage was supposed to be zero hours. Seagate indicate that these disks have a MTBF of 1.2M hours, so 31,000 hours doesn't appear to bethat bad.

Rob

Well they definitely aren't "new and unused", and 31.7K hours is about 3 1/2 years, which isn't bad. The site I'm at today has a NetApp array that has yet to throw a disk, not saying it won't at some point, but here's an example of it's hours...

Code:

Disk: 0a.00.23
Shelf: 0
Bay: 23
Serial: KWG6LBXR
Vendor: NETAPP
Model: X422_HCOBE600A10
Rev: NA01
RPM: 10000
WWN: 5:000:cca056:0bfcf4
UID: 5000CCA0:560BFCF4:00000000:00000000:00000000:00000000:00000000:00000000:00000000:00000000
Downrev: no
Pri Port: A
Sec Name: 0b.00.23
Sec Port: B
Power-on Hours: 42508

If you are going to keep them I know you can format them to 512 bytes with sg3_utils in centOS.

 

[jgreco]

Seagate indicate that these disks have a MTBF of 1.2M hours, so 31,000 hours doesn't appear to bethat bad.

That's not even close to what MTBF means. You can reasonably hope for an enterprise drive to survive ~50-100K hours.

 

[HoneyBadger]

The disks were sold to me as "new and unused"

Pardon my language, but this should be you, to the seller, right now:

SMART stats could potentially be faked, but it would take a special kind of stupid to fake the drives into a worse condition.

Get your money back for sure; these are very used EMC SAN drives (CLAR300 in the model gives it away - they're from an old Clariion) - although that doesn't necessarily mean "useless" - it just means they aren't new drives or worth a new-drive price. At around twenty bucks each, that's more appropriate.

 

[Mlovelace]

Pardon my language, but this should be you, to the seller, right now:
View attachment 30220
SMART stats could potentially be faked, but it would take a special kind of stupid to fake the drives into a worse condition.

Get your money back for sure; these are very used EMC SAN drives (CLAR300 in the model gives it away - they're from an old Clariion) - although that doesn't necessarily mean "useless" - it just means they aren't new drives or worth a new-drive price. At around twenty bucks each, that's more appropriate.

I figured it was either NetApp or EMC originally, though I see way more NetApp drives being resold these days. I've also seen some 528 byte logical block sas drives, but I believe they were from a HPE SAN.

 

[rwillett]

I contacted the company who sold me the disks. They expressed surprise as they claimed they were told they were new. They have told me to return them.

I'll send them back and get all my money back.

Oh well, back to looking for 6-8 x 3-4TB drives, If anybody knows of any in the UK please let me know. I am slowly losing the will to live on this :)

Thanks for all the help. It's a public holiday in the UK , so I'm going to enjoy the sunshine.

Rob

 

[jgreco]

Some sellers don't seem to understand that these devices are instrumented. I had this happen with a bunch of flash drives, "refurbished and professionally tested", where the stats made it clear that the devices had never been tested.

 

[rwillett]

I agree. I was checking on eBay for new drives. They are listed as 'New' but the description is simply 'looks like new. Grade A pull'.

What does that mean? I've never seen a disk drive that looks damaged and I've seen thousands, I suspect other people here have seen 100,000's. They all look like new, even though they could have 100,000 hours on them.

I've also seen 'New (refurbished)', the disks have been 'professionally wiped' as if that makes a massive difference.

I'm giving up on eBay as I can't trust any of them. Will need to find a non-Amazon retailer at a sensible price.

Rob

 

[rwillett]

The upside is I did manage to flash the H310 card WITH a bios into IT-Mode and learnt more about how it all works.

Rob