On Tue, 22 Jul 2008, Sumit Gupta wrote:

> I have to also note here that most of this is done as a part of our 
> performance regression testing, the purpose of which is to make sure 
> that any bugfixes we make don't regress the performance. There is no 
> formal performance data yet.

These results are very encouraging.  I am sure that there is 
considerable performance tuning to be done.  I watched your various 
videos on the web site and it seemed that issues were being 
considered and well-addressed.

Given past zfs history, the impact caused by synchronous-writes / 
cache-flush as compared with a traditional NVRAM-backed RAID array 
would be very interesting to hear about.

Another interesting thing to hear about is how well legacy filesystems 
(which usually use small blocks) will perform on top of ZFS.  ZFS can 
coalese adjacent data into larger blocks but cache flush requests may 
hinder that.

There is nothing to prevent COMSTAR + ZFS from using NVRAM-backed RAID 
arrays on the back end.

Bob
======================================
Bob Friesenhahn
[EMAIL PROTECTED], http://www.simplesystems.org/users/bfriesen/
GraphicsMagick Maintainer,    http://www.GraphicsMagick.org/

_______________________________________________
storage-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/storage-discuss

Reply via email to