Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 06 Nov 2017 20:23:49 +0000
From:      Frank Leonhardt <frank2@fjl.co.uk>
To:        freebsd-hardware@freebsd.org
Subject:   Re: Do I need SAS drives?..
Message-ID:  <5A00C4D5.2010205@fjl.co.uk>
In-Reply-To: <cd7ab44859b519b9f78fa84cb43ef4ff@vvelox.net>
References:  <4DFBCE11-913A-4FC9-937D-463B4D49816C@aldan.algebra.com> <CAOtMX2jeUbSm535Zvd_7aHfQao-dMs5zbU0o3GRWk%2BcmW1Nq=g@mail.gmail.com> <cd7ab44859b519b9f78fa84cb43ef4ff@vvelox.net>

next in thread | previous in thread | raw e-mail | index | archive | help


On 06/11/2017 10:09, Zane C. B-H. wrote:
> In my years of doing decade plus of DC work, I've seen both SAS and SATA
> drives flake and render systems in operable till the offending drive is
> removed.
>

My experience too.
> For Supermicro it will vary between backplanes.
>
Very true indeed. If they go on or off from time to time, that's good 
enough.

>> I'm guessing that you don't have an expander (since you only have 8
>> slots), so item 1 doesn't matter to you.  I'll guess that item 3
>> doesn't matter either, or you wouldn't have asked this question.  Item
>> 5 can be dealt with simply by buying the higher end SATA drives.  So
>> item 6 is really the most important.  If this system needs to have
>> very high uptime and consistent bandwidth, or if it will be difficult
>> to access for maintenance, then you probably want to use SAS drives.
>> If not, then you can save some money by using SATA.  Hope that helps.
>
> Actually most boxes with more than 4 slots tend to be use multipliers.
>
I'm more mixed on that. There are quite a few Dells with eight or 
twelve-slot backplanes, even if it means two HBAs. Apart from better 
performance, the cost of 2xHBA+backplane is bizarrely less than 
1xHBA+Expander. All the Supermicros I've seen have had expanders though.

> As to uptime, that is trivial to achieve with both.
>
> With both it is of importance of drive monitoring and regular self tests.

WHS! Biggest cause of problems is discovering a flaky drive or two AFTER 
the redundant one has failed. I don't know what anyone else thinks, but 
I'm inclined to do a straightforward read of a block device rather than 
a ZFS scrub because (a) I think it's quicker, especially when there's 
not much workload; and (b) it also reads unused blocks, which are 
probably the majority. "Best Practice" says you should do a scrub every 
three months - seems way to long a gap for my liking.





Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?5A00C4D5.2010205>