It depends a lot on what the write pattern is. If the writes a small,
sequential and the drive cache is enabled, I suspect most sata drives
will do a lot more than 120 iops. The exact number would depend on the
drive throughput at a given request size. If the drive has a very large
cache, it you happen to send it a very short burst of small requests
that fit into it's cache, you'd see a very, very high number.
In general the 80-120 iops figure is for random IO, usually measured
with somewhere between 2k and 16k requests. It includes seeks and
rotational latency, neither of which will come into play on a sequential
stream.
Regards,
Tristan
On 2/10/2009 6:32 AM, Bruno Sousa wrote:
As fair as i know SATA disks should be able to deliver around 120 IOPS...
Bruno
Mika Borner wrote:
Bruno Sousa wrote:
If you use SSD, they are the "/guilty/" ones for the high IOPS ;)
Yes, but there are only two of them and mirrored. Despite that,
writes still come from the ARC.
I'm still interested what the theoretical IOPS limit for 1 TB disks
is....
_______________________________________________
storage-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/storage-discuss
