Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


Is 650 days much on harddrives in a server?
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

Is 650 days much on harddrives in a server?

As some has seen in other threads, I got me a Hetzner Intel Core i7-3770, 32 GB RAM, 2 x 3 TB server via serverbidding yesterday. When I check S.M.A.R.T I can see that the disk (an most likely the server) has been online for around 650 days. They are using Toshiba drives.

Is this much for a server? Personally I have disks that has been on for over 2000 days on one of my home servers (just replaced now) and there has been no issues at all. My experience is that if a disk fails, it fails within a year, if not, it almost never fails. (I have around 30 disks running 24/7/365 at home).

What do you think? Whats normal for second hand servers? I can see at on my Kimsufi KS-3 server the disk has only been on for 66 days now. So most likely the server is just 66 days old, or they have replaced the disk.

«1

Comments

  • Depends what disk it is. The enterprise HDDs come with 5 years warranty. Just age is not a good enough indicator of a disk failing. Of course if you get some 5-6 years old HDD you should be worried, but 650 days is not that much.

  • rm_rm_ IPv6 Advocate, Veteran

    said: I have disks that has been on for over 2000 days on one of my home servers (just replaced now) and there has been no issues at all. My experience is that if a disk fails, it fails within a year, if not, it almost never fails.

    You have answered yourself.

    said: What do you think? Whats normal for second hand servers?

    Do you see them tossing a perfectly working drive into the trash bin, just because it got more than X days running? And do that even for their low-margin cheap dedi lineup (robot/auction)?

    Thanked by 1netomx
  • @rm_ said:
    Do you see them tossing a perfectly working drive into the trash bin, just because it got more than X days running? And do that even for their low-margin cheap dedi lineup (robot/auction)?

    no, I don't. When do you actually get new servers? I understand that with serverbidding you always get second hand servers, they inform very good about that.
    But if I ordered a normal Hetzner, Kimsufi, Online server, do I get a brand new one, or do they resell old servers there also?

  • Most Kidechiere have had their disk online for about 4 years and still fine!

    Thanked by 1myhken
  • LeeLee Veteran

    myhken said: But if I ordered a normal Hetzner, Kimsufi, Online server, do I get a brand new one

    Nope, you might get one with a lot more hours. A "normal" server does not come with a new drive every time it is sold and resold. In fact you could find that one of the drives has 2000 hours and the other is new with 300 hours.

    The chances of you getting a brand new server from anyone is pretty slim.

    Thanked by 1myhken
  • I always got brand new HDD for my SYS server.

  • rm_rm_ IPv6 Advocate, Veteran

    edan said: I always got brand new HDD for my SYS server.

    Perhaps SYS is such a good deal (and the setup fee works), that very few people cancel their servers. :)

  • (05) and (C5) might be the only thing I would concern, unless the HDD is really old, say, 5+ years.

    And IIRC Toshiba HDD rocks.

  • Not everyone gets a server with new HDD.

  • LeeLee Veteran

    I see people on WHT especially, complaining they got a "new" server with a few thousand hours on the drives and the provider refuses to refund their money or replace the drives.

    Honestly don't know where many peoples expectations come from these days.

    Thanked by 1ucxo
  • 650 is more than fine. still highly reliable

  • http://puu.sh/fh7Tv/590fd1f805.png My backup hdd, not in a server but still... its very old and its seen a lot of action. So I guess 650 days is not that big of a deal.

  • @Makkesk8 said:
    http://puu.sh/fh7Tv/590fd1f805.png My backup hdd, not in a server but still... its very old and its seen a lot of action. So I guess 650 days is not that big of a deal.

    I see 63 hours in that image?

  • @kcaj said:
    I see 63 hours in that image?

    WTF its 27571 hours :P

  • I read that most hard drives fail within the first two years (was it an article by google?) and that hard drives that last this long possibly last "forever". So maybe 2 years age is not a bad thing?

    Thanked by 1myhken
  • J1021J1021 Member
    edited January 2015

    @TinyTunnel_Tom said:
    WTF its 27571 hours :P

    Ah, I was reading the raw output at the bottom of the inage which differs to the results at the top.

  • @kcaj said:
    Ah, I was reading the raw output at the bottom of the inage which differs to the results at the top.

    hehe :P

  • i got some old drives in a dacentec server and they are reliable (although not speedy)

    http://puu.sh/fhcDD/0f54df0fe0.png

  • qpsqps Member, Host Rep

    We have many disks with over 50,000 hours that are still running like the day they came out of the box from the factory. Just because a disk is old does not mean it is going to stop working. If you monitor a disk properly, it will show signs of failure long before it actually fails in most cases (such as SMART errors).

    In our experience, the most common failure is within the few months of use. If a drive makes it past that, it usually lasts a long time.

    Thanked by 1Maounique
  • CharlesACharlesA Member
    edited January 2015

    Pretty much echoing everyone else where. I've got a few drives with 30K+ hours on them of running 24/7 that haven't had issues yet. One or two started throwing bad sectors, but those were replaced and the others are still running strong.

    I've had drives be DoA or dead after a few weeks or months, but if they don't give you any problems that'll last for a pretty long time.

    Just be sure you have a backup plan in the event the drive or drives do go out.

    I just checked my main server and all four drives in my storage array have ~13393 hours on them, which is about a year and a half, which was when I upgraded that drives in that guy.

  • @CharlesA said:

    Just be sure you have a backup plan in the event the drive or drives do go out.

    Of course, this server will be used as a backup server for my personal files, that I'm only loose if the house burns down, or someone takes all my computers (and then I mean all of them). And I have a Norwegian online backup service for the same files, so that has to go also. First then will I need to download the backup from my new server.

    I will also use it for one of my live backup servers, supporting my main servers, with DNS fail over. But again, it will be one of three servers, so first my main server has to go, so my second live backup server (of course with another provider then my main provider) has to go, then I need this server to be online.

    I will also use it for storing Virtualmin backup files on from my main servers. Again, just one of many servers there I store backup files, taken almost every day each week.
    Downloaded to my home server, saved for a long time, also stored on my online backup service. So hell will freeze to ice, before I really need the backup files on the server.

    But still, it will be a part of my setup, playing a good role in my total mix of VPS servers and now a dedicated server. And it gives me lots of extra disk space I can use for storing something not very impotent. I can use it to create VMs with Hyper-V (two are running now, one Linux server for my live backup, and one Windows for FTP server).
    I still have one more IP I can use on a new VM.

    And then I don't use the main server to anything. (or it runs VMs). All that for €30/mo. Not a bad deal.

  • @qps said:
    We have many disks with over 50,000 hours that are still running like the day they came out of the box from the factory. Just because a disk is old does not mean it is going to stop working. If you monitor a disk properly, it will show signs of failure long before it actually fails in most cases (such as SMART errors).

    In our experience, the most common failure is within the few months of use. If a drive makes it past that, it usually lasts a long time.

    Agreed. I justed checked my HDD which was hooked up to my PC (before I had my RPI) and found around 40k hours on it. Not failed once, no bad sectors.

  • myhkenmyhken Member
    edited January 2015

    I have checked now with CrystalDisk (that I saw others here using) and here is the results:

    Seems OK for me, do you agree?

    PS! Seems like they have stayed on for a long time, with very little on and off if you see the Power on count. I see others has ALLOT higher numbers there.

  • myhken said: Seems OK for me, do you agree?

    seems perfect!

    Thanked by 1myhken
  • What bullshit is this? You got a server through serverbidding which indicates by itself that this is a used server. But of course LET idiots expecting brand new hardware at the price of a used server.

  • myhkenmyhken Member
    edited January 2015

    @Amfy said:
    What bullshit is this? You got a server through serverbidding which indicates by itself that this is a used server. But of course LET idiots expecting brand new hardware at the price of a used server.

    Read my posts, i have been clear on that of course I knew that I got used hardware. Too that price I did not expect new hardware. I only asked if the disk was "too old" or it's normal age on them, server vice. So I do not hope that it's was me you was talking about.

    http://lowendtalk.com/discussion/comment/888621/#Comment_888621

    I understand that with serverbidding you always get second hand servers, they inform very good about that.

  • nexmarknexmark Member
    edited January 2015

    Toshiba, Or Hitachi or Seagate... I don't feel good with 2+ years on it




    But Western Digital, I do









    Here's my Kimsufi




    I've never seen a Hitachi fail on me yet, so they must be good, however this HDD I got was kinda new when I got it (238) hours

  • There's not that many factors where providers can make some margin. Probably the most used approach, I guess, is to get good second-hand hardware from a good source (preferably yourself, your premier services department).

    That said, I do know providers who don't like to use 2nd-hand drives. And anyway, disks for VPS are virtually always RAIDed (and now we know one more reason ...).

    Thanked by 1myhken
  • If the drive isn't erroring then it shouldn't be a problem and if it is showing errors then show them evidence of this and they should replace it for you.

    I had one of the serverbidding machines with a faulty drive in (The software raid1 couldn't even sync properly) I sent them the errors from the logs alongside a smartctrl and it was replaced within about 15mins of raising the ticket.

    As long as you have backups then it's not world ending if a drive did die, just inconvenient

    Thanked by 1myhken
  • The primary drive on my Windows server has just under 30,000 hours on it. Given that there is not much I/O on this drive, the hour count is fine. Even if the drive were to die, I have a drive image of it and can be up and running on a new drive in a very short time. The other drives listed ( E: F: H;) all have 14,000 to 15,000 hours on them and no concerns about them failing.

Sign In or Register to comment.