Which NAS drive would you get?

Discussion in 'SSDs & Data Storage' started by LFaWolf, Dec 6, 2017.

  1. LFaWolf

    LFaWolf Limp Gawd

    Messages:
    332
    Joined:
    Aug 7, 2016
    Last edited: Dec 6, 2017
  2. sinisterDei

    sinisterDei Gawd

    Messages:
    671
    Joined:
    Dec 1, 2004
    I prefer the HGST. Their drives have lower overall failure rates on Backblaze's drive reliability reports. It's not enough of a difference to *pay extra* for it, but if they're equal price - heck yeah get the HGST.
     
  3. LFaWolf

    LFaWolf Limp Gawd

    Messages:
    332
    Joined:
    Aug 7, 2016
    I saw the Backblaze failure rate of the Seagate drives as well, but those are not the Ironwolf but perhaps the even better Enterprise drives. Do you think the manufacturing process is the same too? I am actually leaning toward the Seagate but the failure rate is quite high, but of course they use a lot of the Seagate drives.
     
  4. Prisicilla

    Prisicilla n00bie

    Messages:
    38
    Joined:
    Nov 15, 2017
    I would pick HGST 10TB He drives, if price wasn't a factor.
     
  5. sinisterDei

    sinisterDei Gawd

    Messages:
    671
    Joined:
    Dec 1, 2004
    The Seagate failure rates aren't particularly high. They are higher than the HGST when comparing their 4TB units, but Backblaze hasn't bought enough 8TB+ HGST drives to make a valid comparison between the two brands. BB has used almost exclusively Seagate brand drives when purchasing 6TB+ drives, but they've discussed that in the past and it is a matter of pricing and availability to them, not reliability.

    Even when comparing the HGST and Seagate drives at 4TB, where the HGSTs have a lead, the Seagates are still not bad. As of 2017, they have a ~3% annual failure rate on the Seagates and a ~0.5% annual failure rate on the HGSTs. It's better on the HGST for sure, but when you're talking about an array with only 4 drives in it even a single drive failure for either brand would be a statistical anomaly outside of the norms set by the observed failure rates. Plus, those are all for drives that are a few years old, and whose failure rate is likely increasing by the year (the ST4000DM000 was released in late 2012/early 2013 I believe).

    What it comes down to for me is this; the HGST has better average observed failure rates, but only at a scale that none of us here on this forum are likely to witness. Thus, I choose the HGST when all other factors (price) are equal; I'm not willing to pay more $$$ for a statistically negligible lower chance of a drive failure, but if I can get that negligibly lower chance of failure for free I'll do it.
     
    LFaWolf and drescherjm like this.
  6. bigblueshock

    bigblueshock [H]Lite

    Messages:
    70
    Joined:
    Oct 27, 2017
    I have a Seagate Ironwolf 8TB. No issues after a year so far (knock on wood)

    HGST if money isn't an issue.

    Not a fan of Western Digital's speed/benchmarks.
     
  7. LFaWolf

    LFaWolf Limp Gawd

    Messages:
    332
    Joined:
    Aug 7, 2016
    I guess it is overwhelmingly in favor of the HGST drive. Thanks everyone!
     
  8. SamirD

    SamirD Gawd

    Messages:
    802
    Joined:
    Mar 22, 2015
    We use both WD RE-series and the HGST. We've had a WD RE or two fail inside of warranty, and no issues on the HGST (knock on wood).

    Of course, WD now owns HGST so the storage wars have come down to Seagate and WDC in terms of traditional hard drives.
     
  9. Frank Bello

    Frank Bello n00bie

    Messages:
    8
    Joined:
    Dec 27, 2016
    Just bought 2 x HGST 10TB He and I was pleasantly surprised that these run both cool (cooler than the HGST 4TB drives they replaced) and quiet. BlackMagic gives the Raid0 of the two drives 390MB/s read and 378MB/s write.
     
  10. SamirD

    SamirD Gawd

    Messages:
    802
    Joined:
    Mar 22, 2015
    How long did it take to transfer all your data to the new 10TBs?
     
  11. Frank Bello

    Frank Bello n00bie

    Messages:
    8
    Joined:
    Dec 27, 2016
    A couple of days, having said which. the array isn't full.

    12TB in 2 days - so about 70MB/sec. It's a puzzler to know what the bottleneck is in the copy process - it's not the raw transfer rate of the disks, nor the thunderbolt2 cable on the back of the enclosure, nor CPU/memory on the Mac mini. Those resources are consistently not saturated. That only really leaves inefficiencies in the operating system and the file structure on disk.