Re: iSCSI: Towards Urgent Pointer Consensus

[Matt Wakeley <matt_wakeley@agilent.com>]

Venkat Rangan wrote:

> In cases 3) and 4) also, we do not need Urgent Pointer, because
> the bandwidth achieved will drop to a point (in response to packet
> drops), where you do not need a large amount of anonymous buffers.
> Additionally, the receiving end in both cases is a custom TCP
> implementation. It is possible to design such implementations
> where anonymously held buffers get "transferred" to its rightful
> owner when sequence holes get filled, without any additional data
> copies.

If the "owner" allocated a buffer for the data, and the data gets put into
"anonymous" buffers, then you can't just "transfer" the buffer to the owner. 
The "owner" already has a place for it, and expects it to be put there.  So
you *will* have data copies.

-Matt

> As long as the TOE memory address space is also addressable
> by higher-level (iSCSI) software layer, this copy can be avoided.
> Perhaps, your "custom implementation" assumes certain constraints
> on memory addressability between iSCSI and TCP layers.
> 
> Venkat Rangan
> Rhapsody Networks Inc.
> www.rhapsodynetworks.com