On Thu, Jun 18, 2009 at 2:14 PM, FUJITA
Tomonori<fujita.tomon...@lab.ntt.co.jp> wrote:
> On Thu, 18 Jun 2009 13:33:42 +0200
> Bart Van Assche <bart.vanass...@gmail.com> wrote:
>> Numbers please. From the last numbers I have seen the latency
>> difference between 10 GbE and fast non-Ethernet networks is still
>> significant.
> I think that 10GbE vendors are happy to give you lots of results.
> Maybe it's still significant for you but it's not for some, which I
> had worked with.

Thanks for the reference. The cited publication mentions a latency of
10.49 µs for an advanced 10 GbE interface. This is relatively high
compared to the latency of e.g. InfiniBand (between 2 and 3 µs).
Whether or not this difference matters depends on the workload. As
known for filesystem I/O the block sizes between 4 KB and 64 KB
dominate. These block sizes correspond to a transfer time of 3 µs and
49 µs respectively at a rate of 10 Gb/s. So for a workload that
consists mainly of 4 KB blocks, the latency difference is important
but for a workload that consists mainly of 64 KB blocks, the latency
difference is not that important.


You received this message because you are subscribed to the Google Groups 
"open-iscsi" group.
To post to this group, send email to open-iscsi@googlegroups.com
To unsubscribe from this group, send email to 
For more options, visit this group at http://groups.google.com/group/open-iscsi

Reply via email to