On Wed, Mar 5, 2014 at 8:34 PM, Greg Zartman <[email protected]>wrote:

> Yep, I get that.  I was referring to mixing vdevs with different drive
> types(Manf) and sizes in the zpool.  One of the technical aspects that
> someone over on FreeNAS pointed out to me is that ZFS does some interesting
> things with spindle speeds and some other internal drive mechanisms to sort
> of "equalize" the zpool.  I wish I could find the thread, but can't.
> Bottom line is this developer was stating that ZFS isn't as sensitive to
> mixed drives types (Manf) and sizes as say a hardware raid solution.
>

Wouldn't this also depend on both the number of platters and data density?
  The longer you data can stay the outer rim, the faster your seq
performance will be.

I've never really tested that with zfs, but is usually to see with Linux md
raid - reduced performance for data closer to the spindle.

It would seem quite complicated for zfs to track this level of specific
hardware knowledge, unless it did it on the fly.

Nicholas



-------------------------------------------
smartos-discuss
Archives: https://www.listbox.com/member/archive/184463/=now
RSS Feed: https://www.listbox.com/member/archive/rss/184463/25769125-55cfbc00
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=25769125&id_secret=25769125-7688e9fb
Powered by Listbox: http://www.listbox.com

Reply via email to