Which hard drive?

    • Offizieller Beitrag

    Statistics are interesting once sample size matches. If your sample size is way lower the numbers others collected are simply irrelevant.

    All I have to say is false. Sample size is only relevant when generating the statistics. As long as the sample size is large enough (which it is), the statistics are relevant. You are referring to luck.

    omv 7.0.4-2 sandworm | 64 bit | 6.5 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.10 | compose 7.1.2 | k8s 7.0-6 | cputemp 7.0 | mergerfs 7.0.3


    omv-extras.org plugins source code and issue tracker - github


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    • Offizieller Beitrag

    I never thought there would be such a disagreement with good published statistics. I will just STFU since my suggestion sucks and everyone has better ideas...

    I agree with you.


    Data is data and while it's not absolute, BackBlaze's data set is massive. It's quite possibly the largest study of hard drives in the world. Of the drive models specified, their failure data is an indicator of what might be expected. Sure, one could buy an exact model where BackBlaze's failure rate was at or close 0.0% (the WD or Toshiba models for example) and it might fail quickly in a desktop. That's the nature of statistics and probability, "nothing is absolute". However, their data is a strong indicator of the probability of failure, at least where the models in their data set are concerned.


    Further, there are reasons why BackBlaze is compiling this data and it's certainly not for the computing community. This data is being compiled, analyzed and actively applied to save money, period. Toward that end, in reducing tech labor hours, down time, etc., it's an active effort toward buying drives that last longer, even marginally, and influence OEM drive prices. It's a matter of trade offs related to cost. Otherwise, compiling such a massive data set would be an expensive and foolish waste of time. (If that was the case, being so poorly managed, BackBlaze would be out of business.)


    While It's possible to poke holes in the collection techniques behind any data set, given the content of this thread, it also appears that there's a great deal of variation in individual interpretation of results. :)
    _________________________________


    In balancing out the picture, a quote from Mark Twain comes to mind:
    "There are lies, damned lies, and statistics."

  • one could buy an exact model where BackBlaze's failure rate was at or close 0.0% (the WD or Toshiba models for example) and it might fail quickly.

    Exactly. Since the single disk you bought is not affected by Backblaze's statistics. It can fail right now or it can fail in 10 years. The only certainty involved is that it will fail for sure. It's only a matter of time when this will happen and a matter of strategy how to deal with this (100% disk capacity wasted for RAID-1 being the worst choice of course)


    You are referring to luck.

    This whole thread is about betting on luck and 'a good feeling since $something done wrt redundancy' ;)

  • I read everything until here, now my copy&paste from another thread of mine because I'm still not sure.


    Do we really need NAS drives? Why not Desktop drives?


    Zitat

    In some days or weeks I would like to buy a new HDD for my NAS build.
    Curently there is a 2 TB WD Red and Seagate 2 TB (FireCuda - wrong delivery! Bought IronWolf normally) inside.


    These 2 TB aren't enought, this is for sure. So I would like to buy 4 TB drives next.


    But which one? Western Digital Red, Seagate IronWolf, something else?

  • NAS drives have optimized firmware,that's all. Especially for RAID recovery, but if you use ZFS,then i guess it doesnt add much though.
    But i have had good experience with RAID drives in las 7-8 years so i always recommend them.

  • I do not use RAID or ZFS.


    I prefere the old school way by using rsync once a day or a week and then I backup everything to an external drive. And I use ext4. ZFS is way too complicated for me.
    I don't need 100% availability. If a drive fails, it fails.


    So NAS drives just have optimized firmware? And because of that they cost so much more? I mean... there is a Seagate Barracuda 4 TB for around 60€ out there.
    Why not using this?

    • Offizieller Beitrag

    So NAS drives just have optimized firmware? And because of that they cost so much more? I mean... there is a Seagate Barracuda 4 TB for around 60€ out there.Why not using this?

    Setting aside obvious differences like rotation speed and form factor; mechanically, there's very little in the way of real differences between spinning hard drives. The media itself, the platters, are made by only a hand full of manufactures world wide. Drive OEM's, from time to time, will make an incremental improvement. If it proves to be a useful improvement and easy to implement, in a year or two, all the major drive OEM's rapidly adopt the change.


    With that said, optimized firmware can make a real difference in specific applications like security DVR's were drives are writing continuously, 24x7.


    However, Backblaze.com believes as you do, that consumer commodity (read "low cost") drives do the job and they use them, by the thousands, in large storage arrays. While they didn't specifically state it this way, for Backblaze, it's about "cost versus longevity". I'm in this camp as well. Most home NAS systems are not running their hard drives continuously. They don't need blazingly fast, high (rotation) speed drives, with firmware optimized for data centers. Low cost consumer desktop drives will do the job. I actually prefer 5400rpm drives because they consume less power and, with the slower rotation speed, run a bit cooler.



    But, that's just an opinion. Others may have differ.

  • The price between non-NAS and NAS drives isn't that high.
    Western Digital Blue 4 TB EZRZ 100€, Western Digital Red 4 TB EFRX 110€ to 125€.


    The red is 5400 RPM, the Blue 7200. I think i will stick with the Red ones.
    I already have a 2 TB RED so they can greatly work together.


    Seagate runs with 5900 RPM. I would buy Seagate too but everyone says they fail.



    To be honest: I really don't know what to buy.
    Normal Desktop HDD (e.g. Seagate BarraCuda 4 TB which runs 5400 RPM) or NAS drive... I don't do anything regarding RAID.

    • Offizieller Beitrag

    From the Blackblaze study, most of the WD, Toshiba, and NGST models tested seemed to have longer life when compared to other brands.


    Yeah, Seagate does get a bad rap. However, I'm, running 2 Seagates at 5900rpm in a 4TB ZFS zmirror. They're desktop drives. At 10K hours, so far, they've been fine.
    On a backup server, I'm running 2 Toshiba's (NAS drives I got on sale), at 7200rpm , in a zmirror and I'm Rsync'ing the mirrors' entire contents to a 4TB UnionFS drive. That gives me two full copies of data on the same backup server. And I have another full copy on an SBC, with a single 4TB drive (for 4 copies total). With personal data that goes back to Windows 3.1, I'm taking no chances. :)


    What it boils down to is, with checked and tested full copies of your data (rock solid backup), you won't have to worry about a single drive failure (which is inevitable). Solid backup also frees you up to buy what you want, without risk. I'm not worried about a single drive failure, but the cost of setting up is a consideration. In any case, I'd rather spend money on on 2 cheap drives and have 2 copies, than buy one high end drive and have one copy.


    BTW:
    I'd doing zmirrors for bitrot protection, not availability. I see a 2 drive mirror as a single drive because, functionally, that's what it is. (If one drive fails, there's a real risk that the other drive will be corrupted.)

  • Ok I think it will be the Seagate Barracuda 4 TB ST4000DM004 ones.
    They are OK in price, often in sale and 4 TB. It's 20€ cheaper than the Seagate IronWolf 4 TB.


    But the best is, they cost low in electricity. Less than a NAS drive.

  • Backblaze study isn't useful to normal users,mainly because the cramp up 40 of those drives, so this is use case you will never make.
    Take the ones that are available to you,Seagate or WD doesn't matter.

  • Personally, I am a fan of buying these drives from Best Buy and shucking them out of their enclosures. They are 8 TB WD Red drives inside (ok, they may be slightly different). They are very frequently on sale for $139-$149, and for me, that's cheap enough that I am not too concerned about how they compare to other 8TB drives. There are other brands/models I would prefer, but the price/capacity is so low that I just can't resist.

  • I have 4 seagate 5900 and 2 7200 2tb and 3 4tb. All the 5900 are over 30k hrs. The longest is 58k hrs. And I shutdown everynight. Newer 7200 are over 10k too. All of them are still going strong. I must be incredibly lucky.

    • Offizieller Beitrag

    Personally, I am a fan of buying these drives from Best Buy and shucking them out of their enclosures. They are 8 TB WD Red drives inside (ok, they may be slightly different). They are very frequently on sale for $139-$149, and for me, that's cheap enough that I am not too concerned about how they compare to other 8TB drives. There are other brands/models I would prefer, but the price/capacity is so low that I just can't resist.

    An 8TB drive for $140 or so,,, that's a pretty good deal.


    - The warranty would be ended when you open up an external drive so, I'll take it that you've had no problems so far?
    - How many have you opened? (And)
    - Have you ever tried to reuse the enclosure with another hard drive?

    • Offizieller Beitrag

    Which should I choose, cache or RPM?

    It depends on the application. If it's large files, continuous reads (video files, big DB's, etc.), multiple users, that leans toward RPM. If it's small files read over and over (as in an OS), that leans toward a larger cache.

    For home NAS use, I tend to go with a good brand that's on sale. If that includes good spec's, that's a bonus.

    • Offizieller Beitrag

    And about the recent discussion about SMR and CMR is it something that I'll notice on this kind of device?

    I'm not sure what you're asking here. "Noticing a difference" would depending what you're doing with the drive. An SMR (Shingled) drive can pack more data in, but they have a few down sides to consider such as being slower and there's a maintenance mode. SMR's are supposed to be cheaper, but they're fairly rare - the idea didn't take off. A CMR, on sale, can be had for the price of an SMR.

Jetzt mitmachen!

Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!