Seagate 8TB Archive Drive in FreeNAS?

Status
Not open for further replies.

wpirobotbuilder

Dabbler
Joined
Apr 21, 2014
Messages
16
BS.

The URE rate is 10^14 still -> read error at least every 12TB. chances of a mirror rebuild failing are ~65%. I'd only use it in raidz2 or raidz3 for redundancy even during rebuild. My design would be 4x 11disk z3 + 1 spare in a 45drives.com chassis -> >200TiB usable with 2 parity disks per vdev (one parity disk is needed just to be able to read all that data correctly).

Whoops, I didn't see that it has no TLER support. Yeah, mirrored isn't the way to go. Then again, since it has no TLER support, you probably shouldn't be using it in a RAID configuration, period.
 

robo989

Cadet
Joined
May 8, 2015
Messages
6
Thanks for your reply. Would you buy them again, if you had the chance to go back in time? :)

Have you had any issues?

Hi there sorry for late reply...
No issues, been running Freenas 9.3 RAIDz2 now with 8 of these for the past 6 weeks or so.

Here's some rebuild stats I posted on another forum a while ago....


  • 8 x 8TB Seagate Archive/FreeNAS 9.3/ZFS/RAIDz2 (2 drive redundancy)
  • Full rebuild sector by sector on drive 67% full/35TB of reads over entire array/4.39TB writes to resilvered drive.

  • Resilver/rebuild complete in 107 hours - 4.5 Days.

  • See previous post for scrubbing time - 35TB in 12 hours~


  • Quote:
  • pool: Seagate
  • state: ONLINE
  • scan: resilvered 4.39T in 107h51m with 0 errors on Fri May 8 23:14:18 2015
  • config:

  • NAME STATE READ WRITE CKSUM
  • Seagate ONLINE 0 0 0
  • raidz2-0 ONLINE 0 0 0
  • gptid/1c091021-edd7-11e4-b29f-0cc47a324cfc ONLINE 0 0 0
  • gptid/1c8b74f0-edd7-11e4-b29f-0cc47a324cfc ONLINE 0 0 0
  • gptid/1d070db4-edd7-11e4-b29f-0cc47a324cfc ONLINE 0 0 0
  • gptid/1d847228-edd7-11e4-b29f-0cc47a324cfc ONLINE 0 0 0
  • gptid/8d27911b-f247-11e4-a2b8-0cc47a324cfc ONLINE 0 0 0
  • gptid/1e8e6f48-edd7-11e4-b29f-0cc47a324cfc ONLINE 0 0 0
  • gptid/1f228ab9-edd7-11e4-b29f-0cc47a324cfc ONLINE 0 0 0
  • gptid/1fb1ec99-edd7-11e4-b29f-0cc47a324cfc ONLINE 0 0 0

  • errors: No known data errors

  • pool: freenas-boot
  • state: ONLINE
  • scan: none requested
  • config:

  • NAME STATE READ WRITE CKSUM
  • freenas-boot ONLINE 0 0 0
  • ada0p2 ONLINE 0 0 0

  • errors: No known data errors
  • [root@freenas] ~#
Scrubs are super quick mind, between 750-900MB/s
 

robo989

Cadet
Joined
May 8, 2015
Messages
6
Thanks for your reply. Would you buy them again, if you had the chance to go back in time? :)

Have you had any issues?

Definitely provided I was doing at least Z2, over 8 drives I'd go z3.
I wanted the smallest footprint possible so it's on a supermicro itx board in a U-NAS 800 case, nice and small and pretty powerful Xeon-D box.

If I wasn't bothered about physical footprint of server then I'd have just got a server chassis and 4TB drives.

No point in this unless you're after space savings.
 

Lix

Dabbler
Joined
Apr 20, 2015
Messages
27
Thanks for posting again, very tempted to run a few of these in my FreeNAS backup rig.
 

Arwen

MVP
Joined
May 17, 2014
Messages
3,599
robo989,
Glad they worked out for you.

In a mostly read environment, they should work pretty good, as your scrub numbers show.

Based on my limited knowledge of the Seagate Archive disks, they use a small, faster cache,
(which is non-shingled disk space), for writes. Then flush as activity permits, (or timeout
occurs). But, if this write cache is full, (as a re-silver will do), performance will drop.
 
Last edited:

TXAG26

Patron
Joined
Sep 20, 2013
Messages
310
So how are these 8TB Seagate Archive disks shaking out? Some of the original posters have had these now going on 6+ months.

I'm kicking around picking up 3 or 4 of these 8TB drives for a dedicated backup server. Drives would be setup as either RaidZ1 or RaidZ2.
3x 8TB RaidZ1 = 14.6TB (Est. $720)
4x 8TB RaidZ2 = 14.6TB (Est. $960)

Is the lack of TLER a show-stopper for using these with ZFS/RaidZ? I would be writing to these drives maybe a couple times per month when long-term backups run. I'm guessing that running RaidZ2 would lessen any negative effects of not having TLER as I could always just pull the drive that is hanging everything up and still have some redundancy, right?

The other alternative was to pick up two of the USB 3.0 8TB externally cased Archive drives and use those independently. However, being USB external drives, it would probably be difficult to run ZFS on them and even then, they would not correct errors when scrubbed (just recognize the error).

Thoughts?
 

rogerh

Guru
Joined
Apr 18, 2014
Messages
1,111
My understanding was the ZFS coped quite well with lack of TLER. At least, it may stop responding for time, but doesn't usually throw out the disk when it finally comes to life again. The experts will confirm/refute.
 

Ericloewe

Server Wrangler
Moderator
Joined
Feb 15, 2014
Messages
20,175
My understanding was the ZFS coped quite well with lack of TLER. At least, it may stop responding for time, but doesn't usually throw out the disk when it finally comes to life again. The experts will confirm/refute.
Right, but the latter behavior has caused some people a few headaches of varying degrees. KB/s-speed scrubs, for instance.

Not a data security issue, definitely a usability issue. The admin will have to decide if it's worth the extra cash.
 

Arwen

MVP
Joined
May 17, 2014
Messages
3,599
So how are these 8TB Seagate Archive disks shaking out? Some of the original posters have had these now going on 6+ months.

I'm kicking around picking up 3 or 4 of these 8TB drives for a dedicated backup server. Drives would be setup as either RaidZ1 or RaidZ2.
3x 8TB RaidZ1 = 14.6TB (Est. $720)
4x 8TB RaidZ2 = 14.6TB (Est. $960)

Is the lack of TLER a show-stopper for using these with ZFS/RaidZ? I would be writing to these drives maybe a couple times per month when long-term backups run. I'm guessing that running RaidZ2 would lessen any negative effects of not having TLER as I could always just pull the drive that is hanging everything up and still have some redundancy, right?

The other alternative was to pick up two of the USB 3.0 8TB externally cased Archive drives and use those independently. However, being USB external drives, it would probably be difficult to run ZFS on them and even then, they would not correct errors when scrubbed (just recognize the error).

Thoughts?

My 8TB Seagate SMR is working just fine as a single disk ZFS pool for backups. Reads upto 150MBps, and
writes about 30MBps using RSync with whole file copies. Since I have the space, about 1.2TB used in my
FreeNAS at present, I am performing full copies using RSync. Later when the 8TB disk is full, I'll delete the
oldest. When the source get's over 4TB used, then I'll move to RSync incrementals for this backup drive.

Understand that these are slower for continuous writing, (like my 1TB backups). Burst writing is different.
Probably reasonable. If you use differential RSync, which copies only changed data, then it would likely run
faster. Both because it's not copying the same data, AND what data it is copying maybe smaller than the higher
speed non-shingled disk space cache. (I read somewhere that this is 20GBs of space...)

Using RAID-Z1 may be problematic. The comments about drives larger than about 3TB state that you can get
a failed disk read during a replacement disk re-sync process. Thus, causing RAID-5 & RAID-Z1 to fail. So the
general recommendation is to use RAID-6 or RAID-Z2.

I do use mine with an external eSATA enclosure. It's a decent box, with 2 hot swap disk bays, fan and USB 3.0
with UAS as well. This is so I can remove the drive and take it off site, (where it is right now). Natually I don't
use that enclosure's builtin HW RAID...

http://www.amazon.com/gp/product/B00M77UOC8/?tag=ozlp-20

One last comment. I can't use ZFS send and receive. It seems like a useful feature, but I needed my 8TB backup
drive to have ZFS for Linux compatibility, (older ZPool version). Thus, if my FreeNAS melts into a puddle, (or is
stolen), I can access the backup data either from my netbook or miniture media server, both of which run Gentoo
Linux and have ZFS loaded.
 
Last edited:

rogerh

Guru
Joined
Apr 18, 2014
Messages
1,111
I am using FreeNAS's built in replication to replicate to a ZFS on Linux pool and, apart from complaining that it can't set aclmode on the recipient it seems to work ok. I have copied back a few files but not tried a full scale recover, especially not of Time Machine backups as I think something bad might happen if I tried. I am not sure that any method of copying the files is without some potential permission problems.
 

TXAG26

Patron
Joined
Sep 20, 2013
Messages
310
Thanks Arwen and everyone else. I'm going to chew on this a bit more. I might just pick up two of these 8TB external drives and throw NTFS on them and do the off-site backups from a Windows workstation.

It's a shame we're still maxed out at regular 6TB NAS drives, seeing as those were released over a year ago. Prices on the 6TB Reds have barely dropped less than $50 over the past year as well. For consistency-sake, I don't consider the exotic inert gas or laser heater HDD's in the same ballpark as consumer HDD's.
 

Apollo

Wizard
Joined
Jun 13, 2013
Messages
1,449
Just did the plunge on the 8TB drive, namely the Seagate ST8000AS0002 8TB 5900RPM 128MB Cache SATA3 Archive Hard Drive.
So far so good. The main application for this disk is to act as an external single drive backup solution to my Freenas setup.
I am using replication and I can say it is working like a charm.
It is able to write at a sustained 190'ish MB/s and read speed might seem a bit slower but not by much if any.
I was able to perform replication of the equivalent of 5TB of data in less than 12Hrs and at about 5500 snapshots.
The multiple snapshots is what's causing performance issues. The sustained throughput at the end of the replication was still running as high as 145MB/s.
I am currently doing a scrub of the drive to see how it performs.
 

solarisguy

Guru
Joined
Apr 4, 2014
Messages
1,125
Just did the plunge on the 8TB drive, namely the Seagate ST8000AS0002 8TB 5900RPM 128MB Cache SATA3 Archive Hard Drive.
So far so good. The main application for this disk is to act as an external single drive backup solution to my Freenas setup. [...]
How is it connected? External?
 

Apollo

Wizard
Joined
Jun 13, 2013
Messages
1,449
How is it connected? External?
I am connecting it to one of the available SATA ports.
When I am done playing around with it, I will disconnect it for storage.
 

aadje93

Explorer
Joined
Sep 25, 2015
Messages
60
How will these bad boys do it in a 6 disk Z2 vedv? I'm currently in the planning phase of my 24 disk build but for the price (they are cheaper then 6TB wd red!!)

With 4 6disk Z2 vedvs in a pool, 1 vedv only needs to do ~30mb/s read/write to saturate my gigabit connection. For the price i get 128TiB usable capacity instead of ~96TiB, thats a pretty difference for a little less costs, but i'm only reading bad stuff about them from people only running 2-4 disk, i'm maybe going to use 24! of them, so then the "raid" will bring back a little speed?
 

diedrichg

Wizard
Joined
Dec 4, 2012
Messages
1,319
How will these bad boys do it in a 6 disk Z2 vedv? I'm currently in the planning phase of my 24 disk build but for the price (they are cheaper then 6TB wd red!!)

With 4 6disk Z2 vedvs in a pool, 1 vedv only needs to do ~30mb/s read/write to saturate my gigabit connection. For the price i get 128TiB usable capacity instead of ~96TiB, thats a pretty difference for a little less costs, but i'm only reading bad stuff about them from people only running 2-4 disk, i'm maybe going to use 24! of them, so then the "raid" will bring back a little speed?
Aren't the 8TB+ "archive" drives and therefore not intended for traditional active pool storage? I thought I remember some threads about those drives with people trying to use them as you describe and they were getting terrible write (or was it read) performance because... well, that's not what they were intended to do.
 

aadje93

Explorer
Joined
Sep 25, 2015
Messages
60
probarly it was write, but as i'm planning 4 vdevs of z2 6disk arrays (essentially raid 0 on 4 raid 6 arrays) its different then a single array what most ppl do, my build will be a bit "bigger" then average :p
 

Arwen

MVP
Joined
May 17, 2014
Messages
3,599
aadje93,
Yes, it's write performance. The slowest part would be continuous writes that the SMR drive
has to flush from it's non-shingled cache tracks to shingled tracks.

Reads can also be problematic. Since the data is relocated where ever the drive thinks it should
be, what we may think of a simple reads may very well cause head seeks on a SMR drive.

If your use case is read-mostly, and you understand writing is noticably slower, you should be OK.
 

aadje93

Explorer
Joined
Sep 25, 2015
Messages
60
But what kind of writes can i expect? Most users here have 4-8 disk single, or double vdev pools, i'm with 4 vdev pool spread out over 24 disks, the bulk data is very nice :)
 
Status
Not open for further replies.
Top