(Personal rig) Toshiba N300s any good?

[Minxster]

I've seen a small number of posts about these drives but wanted to see what people thinking of them recently? I'm looking at the 6TB drives, specifically:
https://www.ebuyer.com/776380-toshi...ity-nas-hard-drive-at-ebuyer-com-hdwn160uzsva

I'm gearing up to replace half of my array to these 6TB N300s (5 in total out of 12). Heat is an issue but I'm running industrial Noctua fans (push/pull) in my Define XL R2, so I've had no complaints temp vs. tolerance issues so far.

So just seeing if anyone is running a bank or two of these drives? :)

 

[Chris Moore]

I'm gearing up to replace half of my array to these 6TB N300s (5 in total out of 12). Heat is an issue but I'm running industrial Noctua fans (push/pull) in my Define XL R2, so I've had no complaints temp vs. tolerance issues so far.

So just seeing if anyone is running a bank or two of these drives? :)

It is a different model, but I will chime in with my personal Toshiba drive story and you can take it for what it is worth. Around the middle of 2017 I was looking to replace one of the vdevs in my home NAS because the drives in that vdev had been migrated from an even older system and were over five years of age and had been failing with some regularity. It is a six drive vdev that was made of 2 TB drives and I simply wanted newer drives, not to expand capacity. I purchased eight (so I could have spares) Toshiba drives like this:
https://www.newegg.com/Product/Product.aspx?Item=N82E16822149407
They were, at the time, the least expensive 2 TB drives available. I did not expect great things from them but they did exhibit better performance than the drives they were replacing, except for one thing. Within three months, I had three failed drives. That is a higher failure rate than I have ever seen with any other model of drive. It put me right off the Toshiba drives and I went out and bought a different brand of drive. Apparently, these Toshiba drives had already made a name for themselves because I had a very hard time selling the ones that were still working on eBay. I lost about $25 each on them after less than six months. The replacement drives have given me no trouble at all, not even a high fly write.
If you want to give Toshiba a shot, go for it, but I wouldn't. I went with Seagate drives, this is the case (empty now) that I bought:

 

[rvassar]

I have one new Toshiba P300 paired with a tried & true HGST 7K3000 drive @ 27k hours. The P300 is just past 1k hours... I haven't had any trouble... Yet...

Not holding my breath... It's an experiment.

 

[dir_d]

I use only the CA300's 3TB and N300 6TB Toshibas. The 3TB drives have been on 24/7 in my FN 9.3 Machine since 2013 not one drive has died yet. I have 6 but the other vdev of seagate 3TBs have all died but one which i replaced with the CA300 3TBs. The N300s i have now are in my FN 11.1 U4 that has been up since around Jan, no problems though.

 

[Chris Moore]

seagate 3TBs have all died

Seagate 750 GB, 1.5 TB and 3 TB drives have been problem children for years. No surprise there. The Seagate 2 TB and 4 TB drives are significantly better.

 

[pro lamer]

Two months ago my employer bought 3 new PCs with 2TB 7200rpm Toshiba DT01ACA200 and all of them had "mechanical start errors" - count of them was between 10 and 20. One of the HDDs (the one with 20 errors) was RMAed and replaced with the same model and the new one has 6 of them now.

EDIT: Maybe you can try buying various types of disks at once so they will not die all at once since they are all different "lot" (batch) for sure ;)

 

[Minxster]

Cheers for the posts :)... I'll preface this message by saying I'm looking at my large (almost silly) vdev of 12 drives in Z3; I know it's not to a lot of people's tastes but it fits the bill well for what it's intended for. I proactively and periodically buy new larger disks a batch at a time. So the lowest common denominator (size) slowly increases. It takes a bit of forward planning with drive/pool stats along with tracking total size increase year on year. That's where Excel comes in handy :p and it's telling me to step up my game and start scaling again now.

So... the vdev has 3 different drive sizes/types, but for the most part they're WD RED mix of 3TB and 4TB for now... There's a 5TB Tosh x300 in there already as a test, though it did run a little hot initially (45c/47c) but not as hot as I've read online, I also changed the acoustic level to dampen the noise a little :)

I have to take on-board people's comments about Tosh drives, I secretly hoped that the the N300s may have been an overlooked unicorn of a great drive? Newegg reviews do also indicate that they seem prone to failure :-(

Why didn't WD make 7,200 REDs at a somewhat reasonable price?

 

[Johnnie Black]

I have more than a couple of dozen Toshiba drives from the DT01ACA, P300, X300 and N300 series on various servers and they been pretty reliable so far, then tend to get a little hot but IMO no hotter than other 7200rpm drives, and they are very competitively priced.

 

[jgreco]

Well, I replaced three failed Seagate 4TB's with three Toshiba MD04's and they've been fine for several years.

# camcontrol devlist
<NECVMWar VMware IDE CDR10 1.00> at scbus1 target 0 lun 0 (pass0,cd0)
<VMware Virtual disk 1.0> at scbus2 target 0 lun 0 (pass1,da0)
<VMware Virtual disk 1.0> at scbus2 target 1 lun 0 (pass2,da1)
<VMware Virtual disk 1.0> at scbus2 target 2 lun 0 (pass3,da2)
<ATA ST4000DM000-1F21 CC52> at scbus3 target 10 lun 0 (pass4,da3)
<ATA ST4000DM000-1F21 CC51> at scbus3 target 12 lun 0 (pass5,da4)
<ATA ST4000DM000-1F21 CC52> at scbus3 target 13 lun 0 (pass6,da5)
<ATA ST4000DM000-1F21 CC52> at scbus3 target 14 lun 0 (pass7,da6)
<ATA ST4000DM000-1F21 CC52> at scbus3 target 15 lun 0 (pass8,da7)
<ATA ST4000DM000-1F21 CC52> at scbus3 target 16 lun 0 (pass9,da8)
<ATA ST4000DM000-1F21 CC52> at scbus3 target 17 lun 0 (pass10,da9)
<ATA ST4000DM000-1F21 CC51> at scbus3 target 18 lun 0 (pass11,da10)
<ATA ST4000DM000-1F21 CC51> at scbus3 target 19 lun 0 (pass12,da11)
<LSI SAS2X36 0e0b> at scbus3 target 20 lun 0 (pass13,ses0)
<ATA TOSHIBA MD04ACA4 FP2A> at scbus3 target 21 lun 0 (pass14,da12)
<ATA TOSHIBA MD04ACA4 FP2A> at scbus3 target 22 lun 0 (pass15,da13)
<ATA TOSHIBA MD04ACA4 FP2A> at scbus3 target 23 lun 0 (pass16,da14)
# foreach i (12 13 14)
foreach? smartctl -a /dev/da${i} | grep Power_On_Hours
foreach? end
9 Power_On_Hours 0x0032 034 034 000 Old_age Always - 26601
9 Power_On_Hours 0x0032 034 034 000 Old_age Always - 26601
9 Power_On_Hours 0x0032 034 034 000 Old_age Always - 26532


They do run warmer as they're 7200RPM drives.

 

[Chris Moore]

Code:

<ATA ST4000DM000-1F21 CC52>		at scbus3 target 10 lun 0 (pass4,da3)
<ATA ST4000DM000-1F21 CC51>		at scbus3 target 12 lun 0 (pass5,da4)
<ATA ST4000DM000-1F21 CC52>		at scbus3 target 13 lun 0 (pass6,da5)
<ATA ST4000DM000-1F21 CC52>		at scbus3 target 14 lun 0 (pass7,da6)
<ATA ST4000DM000-1F21 CC52>		at scbus3 target 15 lun 0 (pass8,da7)
<ATA ST4000DM000-1F21 CC52>		at scbus3 target 16 lun 0 (pass9,da8)
<ATA ST4000DM000-1F21 CC52>		at scbus3 target 17 lun 0 (pass10,da9)
<ATA ST4000DM000-1F21 CC51>		at scbus3 target 18 lun 0 (pass11,da10)
<ATA ST4000DM000-1F21 CC51>		at scbus3 target 19 lun 0 (pass12,da11)

That ST4000DM000 is the same kind of drive I am using. I have twelve in one pool, six in another (along with six 2 TB drives) and a few spares. I really like that model because they don't throw out as much heat as the 7200 RPM drives and I have not had any failures from them yet.

 

[rvassar]

Why didn't WD make 7,200 REDs at a somewhat reasonable price?

They do... They label them "HGST". :D


(Disclaimer: I am a former employee)

 

[jgreco]

That ST4000DM000 is the same kind of drive I am using. I have twelve in one pool, six in another (along with six 2 TB drives) and a few spares. I really like that model because they don't throw out as much heat as the 7200 RPM drives and I have not had any failures from them yet.

Whereas in my case, they were the ones replaced by Toshibas. Sigh.

 

[Minxster]

They do... They label them "HGST". :D


(Disclaimer: I am a former employee)

ROFL

 

[Minxster]

I have more than a couple of dozen Toshiba drives from the DT01ACA, P300, X300 and N300 series on various servers and they been pretty reliable so far, then tend to get a little hot but IMO no hotter than other 7200rpm drives, and they are very competitively priced.

Good to know... Quite a few reviews seem to be people complaining about failure within a few months, so I'm guessing if you're lucky and outlast that initial time-frame they could be a winner? How long have you been running the N300s Johnnie ?

 

[Minxster]

... Yesterday (lastnight), I sat looking over other drives thinking I should just stay away from Tosh... Today I thought screw it, give it a go, I'm running Z3 so I'm good for three failures at a time... ... As I type this... I need to stop procrastinating; are am I ?... LOL

I might need a rig to do burn-in tests ;)

 

[Johnnie Black]

How long have you been running the N300s Johnnie ?

Not very long, a few months, the X300s over a year, and the others more than that.

 

[pro lamer]

I might need a rig to do burn-in tests ;)

I usually use a laptop with an eSATA port.

Sent from my mobile phone

 

[Minxster]

So I just wanted to post back and say I'm still not too sure what to do overall. I'm wondering about trying the Tosh drives, maybe just do a slow roll-out and see what happens after some burn-in testing.

I've been preoccupied with altering my recordsize settings over my datasets, which has meant cp'ing iSCSI extents and planning no moving my Media, but that's 16TB, so won't be doing a plain cp for that :)

I just want to thank everyone he's posted back on this, for me at least it's been a worth while exercise!

 

[Chris Moore]

If you have a target that is large enough to take the data, you can do a zfs send | zfs receive. It is about the fastest way to get the job done. When I made a copy of my main pool with with 14 TB, it took about 12 hours to copy it from my 12 drive (two vdevs @ RAIDz2) main pool to my 4 drive (one vdev @ RAIDz1) backup pool. When I copied the same data from the primary NAS to the backup NAS, it only took a little over 2 hours, but I had all the drives cabled to the same SAS controller so I didn't need to send the data over the network. The point of that is that the number of drives in the pool significantly impacts the rate that the data can be copied. Going from a 12 drive pool to another 12 drive pool, the transfer happened six times faster than going from a 12 drive pool to a 4 drive pool. The time to copy the data would be increased again if it is being pushed over a network. In the examples I gave, all the drives were connected to the same SAS controller.