[Micronet] Consumer drives shown to be more reliable than enterprise drives

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

[Micronet] Consumer drives shown to be more reliable than enterprise drives

John McChesney-Young
Of possible interest to some on this list:

http://www.computerworld.com/article/2687068/consumer-drives-shown-to-be-more-reliable-than-enterprise-drives.html

I've seen other similar studies that have made me leery of Seagate.

John

--
John McChesney-Young, Administrative Assistant
History of Art Department, 416 Doe MC6020
U. C. Berkeley, Berkeley CA 94720-6020
[hidden email] // voice 1-510-642-5511 // fax 1-510-643-2185

 
-------------------------------------------------------------------------
The following was automatically added to this message by the list server:

To learn more about Micronet, including how to subscribe to or unsubscribe from its mailing list and how to find out about upcoming meetings, please visit the Micronet Web site:

http://micronet.berkeley.edu

Messages you send to this mailing list are public and world-viewable, and the list's archives can be browsed and searched on the Internet.  This means these messages can be viewed by (among others) your bosses, prospective employers, and people who have known you in the past.
Reply | Threaded
Open this post in threaded view
|

Re: [Micronet] Consumer drives shown to be more reliable than enterprise drives

Jon Forrest-3


On 9/29/2014 1:39 PM, John McChesney-Young wrote:
> Of possible interest to some on this list:
>
> http://www.computerworld.com/article/2687068/consumer-drives-shown-to-be-more-reliable-than-enterprise-drives.html

This is a topic I've been studying for years.

The first thing I tried to do was to find out what the
objective differences were between consumer drives and
enterprise drives. That took a while because a lot of
the vendors used meaningless marketing drivel to describe
the differences. But, eventually some differences
emerged.

But, then there was the question of whether these differences
were noticeable in the real world. Of course, there are
several ways to measure this, such as throughput and cost.
But, judging from Backblaze's numbers, these differences
don't result in improved reliability.

If enterprise drives turned out to be more reliable than
consumer drives at the same cost increase, then I suppose
which to buy would be strictly a financial decision. But,
that's not what Backblaze has found.

Cordially,
Jon Forrest
UCB (ret.)

 
-------------------------------------------------------------------------
The following was automatically added to this message by the list server:

To learn more about Micronet, including how to subscribe to or unsubscribe from its mailing list and how to find out about upcoming meetings, please visit the Micronet Web site:

http://micronet.berkeley.edu

Messages you send to this mailing list are public and world-viewable, and the list's archives can be browsed and searched on the Internet.  This means these messages can be viewed by (among others) your bosses, prospective employers, and people who have known you in the past.
Reply | Threaded
Open this post in threaded view
|

Re: [Micronet] Consumer drives shown to be more reliable than enterprise drives

Jonathan Loran

The main reason to buy enterprise drives in an enterprise system is for a better chance of continuity throughout the life of the server.  Enterprise drives tend to have a five year warranty, where as desktop drives are warrantied for shorter periods.  If you have a server (more important in a storage server) with compatibility requirements, it becomes critical you can get replacement drives throughout the lifespan of the system.  Generally, the vendors will replace your failed enterprise drives with compatible parts for the duration of the warranty.  I would never buy and enterprise drive for a desktop system, and likely not for a low IO server.

Jon


-     _____/     _____/      /           - Jonathan Loran -           -
-    /          /           /                IT Officer               -
-  _____  /   _____  /     /     Space Sciences Laboratory, UC Berkeley
-        /          /     /       O: (510) 643-5146 M: (510) 421-2462 -
- ______/    ______/    ______/        [hidden email]        -






On Sep 29, 2014, at 3:41 PM, Jon Forrest <[hidden email]> wrote:



On 9/29/2014 1:39 PM, John McChesney-Young wrote:
Of possible interest to some on this list:

http://www.computerworld.com/article/2687068/consumer-drives-shown-to-be-more-reliable-than-enterprise-drives.html

This is a topic I've been studying for years.

The first thing I tried to do was to find out what the
objective differences were between consumer drives and
enterprise drives. That took a while because a lot of
the vendors used meaningless marketing drivel to describe
the differences. But, eventually some differences
emerged.

But, then there was the question of whether these differences
were noticeable in the real world. Of course, there are
several ways to measure this, such as throughput and cost.
But, judging from Backblaze's numbers, these differences
don't result in improved reliability.

If enterprise drives turned out to be more reliable than
consumer drives at the same cost increase, then I suppose
which to buy would be strictly a financial decision. But,
that's not what Backblaze has found.

Cordially,
Jon Forrest
UCB (ret.)


-------------------------------------------------------------------------
The following was automatically added to this message by the list server:

To learn more about Micronet, including how to subscribe to or unsubscribe from its mailing list and how to find out about upcoming meetings, please visit the Micronet Web site:

http://micronet.berkeley.edu

Messages you send to this mailing list are public and world-viewable, and the list's archives can be browsed and searched on the Internet.  This means these messages can be viewed by (among others) your bosses, prospective employers, and people who have known you in the past.


 
-------------------------------------------------------------------------
The following was automatically added to this message by the list server:

To learn more about Micronet, including how to subscribe to or unsubscribe from its mailing list and how to find out about upcoming meetings, please visit the Micronet Web site:

http://micronet.berkeley.edu

Messages you send to this mailing list are public and world-viewable, and the list's archives can be browsed and searched on the Internet.  This means these messages can be viewed by (among others) your bosses, prospective employers, and people who have known you in the past.
Reply | Threaded
Open this post in threaded view
|

Re: [Micronet] Consumer drives shown to be more reliable than enterprise drives

Jon Forrest-3


On 9/29/2014 5:39 PM, Jonathan Loran wrote:

>   Enterprise drives tend to have a five year warranty, where as desktop
> drives are warrantied for shorter periods.

This is sometimes true, but there have been times when both
types of disks had the same warranty length. But, as the
Backblaze article mentions, the consumer disk drives they
buy come with a warranty that's long enough to cover
the expected usage period of the drive.

Also, even if the enterprise disk drives have longer warranties,
they cost much more than the consumer drive. The point the
Backblaze article makes is that this additional amount is
often excessive.

> If you have a server (more
> important in a storage server) with compatibility requirements, it
> becomes critical you can get replacement drives throughout the lifespan
> of the system.

That's a supply problem, which isn't usually one of the
reasons I've seen for sticking with enterprise drives.
Given that most storage servers have an lifespan
of three years, I'm personally not too worried
about that. Plus, given how many disk drives get
made by each vendor, it would be a very exotic drive
indeed for which you couldn't find a replacement on eBay
or the rest. On the other hand, if you disagree with
this opinion, and you're willing to pay extra, there
are plenty of vendors willing to take your money.


Jon

 
-------------------------------------------------------------------------
The following was automatically added to this message by the list server:

To learn more about Micronet, including how to subscribe to or unsubscribe from its mailing list and how to find out about upcoming meetings, please visit the Micronet Web site:

http://micronet.berkeley.edu

Messages you send to this mailing list are public and world-viewable, and the list's archives can be browsed and searched on the Internet.  This means these messages can be viewed by (among others) your bosses, prospective employers, and people who have known you in the past.
Reply | Threaded
Open this post in threaded view
|

Re: [Micronet] Consumer drives shown to be more reliable than enterprise drives

Jonathan Loran

On Sep 29, 2014, at 6:33 PM, Jon Forrest <[hidden email]> wrote:

>
>
> On 9/29/2014 5:39 PM, Jonathan Loran wrote:
>
>>  Enterprise drives tend to have a five year warranty, where as desktop
>> drives are warrantied for shorter periods.
>
> This is sometimes true, but there have been times when both
> types of disks had the same warranty length. But, as the
> Backblaze article mentions, the consumer disk drives they
> buy come with a warranty that's long enough to cover
> the expected usage period of the drive.

I guess that's really the question, isn't it.  If you are replacing your storage servers every 3 years, then of course, that's easy to come by in desktop drive.  But at our site, we keep our storage servers 5 years.  We don't have the sysadmin bandwidth nor funding to replace every 3.  Of course you could say, buy many spare desktop drives, which you could get quite a few of for the difference in cost between commodity and enterprise drives.  I've never gone that route.

The other part of the equation is I've found in big silos, when your eggs are all in one basket, SAS drives have the edge on performance and stability over SATA.  SAS is a different beast and tend to be enterprise only.  This may not be relevant at all to the type of machine others are using.

Jon


 
-------------------------------------------------------------------------
The following was automatically added to this message by the list server:

To learn more about Micronet, including how to subscribe to or unsubscribe from its mailing list and how to find out about upcoming meetings, please visit the Micronet Web site:

http://micronet.berkeley.edu

Messages you send to this mailing list are public and world-viewable, and the list's archives can be browsed and searched on the Internet.  This means these messages can be viewed by (among others) your bosses, prospective employers, and people who have known you in the past.