|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] RE: iFCP as an IP Storage Work Item> -----Original Message----- > From: Ken Hirata [mailto:Ken.Hirata@Vixel.com] > Sent: Thursday, January 04, 2001 6:57 PM > To: Ips@Ece. Cmu. Edu > Subject: Re: iFCP as an IP Storage Work Item > > > Charles, > > There's one item with your reply that I have a question > about. With regards > to processing overhead, do you agree that FCIP as a tunneling protocol > would not have to look for ELSs in the Fibre Channel > datastream as iFCP > must do? > Hi Ken: First of all, ELS's are well off the performance path, so there's no effect on performance. More to the point, we process ELS's and other frame traffic because doing so gives us a big payoff in cost-performance. A payoff that comes from allowing iFCP implementations to intelligently leverage IP technology. For example, maintaining a seperate session for each N_PORT login gives the gateway a handle for controlling the flows between individual FC storage devices, thus fully exploiting IP-based routing and traffic management. Concealing the Fibre Channel transport infrastructure behind the FC side of the gateway eliminates the need to for a two-plane routing scheme. IP routing is then unconstrained by and fully decoupled from FC routing. More importantly, doing so makes it easier to integrate otherwise incompatible FC infrastructures. While I'm at it, and since this issue has come up several times, I should point out that the NAT-like address translation we do reflects a design choice made to exploit IP scalability. We could have made the tradeoffs differently without affecting iFCP in any fundamental way. Charles Charles Monia Senior Technology Consultant Nishan Systems email: cmonia@nishansystems.com voice: (408) 519-3986 fax: (408) 435-8385
Home Last updated: Tue Sep 04 01:05:58 2001 6315 messages in chronological order |