[mythtv-users] Another take on large-capacity disk enclosures

[Yan Seiner]

On Mon, June 15, 2009 3:08 pm, Brian Wood wrote:
> On Monday 15 June 2009 16:00:36 Yan Seiner wrote:
>
>> Practical considerations:
>>
>> It takes a desktop-class system about 6-10 hours to check a 2 TB raid-5
>> array.  You really don't want a single large array comprising of 20 1.5
>> TB
>> drives; fsck would take weeks.
>>
>> If you pick your hardware right, you can get by much cheaper with esata
>> +
>> port multipliers at little to no performance penalty, especially if you
>> design the wiring so that drives on the same port multiplier are in
>> different logical groups.  Port multipliers come in 1x5 versions, so 4
>> esata ports == 20 drives.  See the sata maillist for hardware that
>> works.
>
> I've heard that claim before, that using port multipliers will have little
> or
> no impact on performance. Is this because today's drives come nowhere
> close
> to the speed of the interface?
>
> Logically it would seem that there is no way to speed up the actual SATA
> port,
> so the theoretical max per port would have to be reduced by using
> multipliers, but if the drives were using far less than the port's
> theoretical speed it might work, so what is the practical limit to
> expansion
> without impacting drive performance?

Questions best asked on the linux-ide list....  From what I gather, it has
to do with NCQ or some such magic, which has to be supported by the esata
hardware, port multiplier, and drives.  I can vouch for that; I have one
esata controller that doesn't do the magic and one that does; the one that
does clocks in as fast (or as slow) as a bare sata drive and the one that
doesn't is far slower.

I "stripe" the raid arrays across controllers, so that a single failure of
cable, multiplier, or esata port will result in a degraded raid rather
than failure.

--Yan

-- 
Yan Seiner, PE