Re: iSCSI Autosense Consensus, Connection next steps

["Randall R. Stewart" <randall@stewart.chicago.il.us>]

Stephen Bailey wrote:
> 
> None of the arguments made for iSCSI connection bundling are unique to
> storage.  If connection bundling is so compelling, TCP or some
> relative will develop its own connection bundling mechanism, and then
> iSCSI can take advantage of it more or less transparently.
> 

Stephen:

About 1 year ago we had this "transport bundling" aka load sharing IN
SCTP. At the request of the IESG it was removed because (if I remember
right) it causes all sorts of problems with Congestion Control and
is still deemed a research topic by the IETF (belonging to the IRTF).It
is aong the same lines (to a small degree) as to why you don't want
multiple TCP
connections.  SCTP now only offers reliablility gained by spanning
multiple
NIC cards on one connection/association but no load balancing. 

Load balancing was looked at as being done (if ever) from an upper
layer, and
in fact there are hooks in the protocol for this... 

A ULP can:

1) A change the primary address (aka where default
   sends go to).. change this every time before 
   sending and you have load balancing

2) Query to get the list of addresses the peer has, useful
   so you know what to change the primary to or override and
   the ULP does not need to maintain some list of peer addresses.

3) Do a directed send that overrides the primary address to
   one of the alternates.

But the actual implemenation of the load balancing is above the
transport layer... :0

R



-- 
Randall R. Stewart
randall@stewart.chicago.il.us or rrs@cisco.com
815-342-5222 (cell) 815-477-2127 (work)