|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: iSCSI Autosense Consensus, Connection next stepsjulian_satran@il.ibm.com wrote: > I am strongly in favor of the asymmetric model but I would like to hear all > possible arguments against it. > > Julo Ok, how about this: The biggest isssue with the "asymmetric" model is that it is NOT "asymmetric" when there is only 1 TCP connection. When there is only one connection, it is the "symmetric" model - I have to do both commands and data on the same connection. Then, when there is more than one TCP connection, I switch over to a different behavior. Does the "asymmetric" model mandate more than 1 connection? (maybe it should) The whole purpose of having multiple TCP connections was to enable some kind of "data stripping". If there are only two links, and one is "low bandwidth", then effectively only one is being used. I suppose that if there are "n" physical links between an initiator/target, that "n+1" TCP connections could be established, with the extra 1 running on one of the connections to transport the commands. -Matt
Home Last updated: Tue Sep 04 01:07:32 2001 6315 messages in chronological order |