The STP was already convoluted. I bet they just didn't think of the additional bells and whistles to speed it up when they were designing it. Or maybe they thought of it but opted for keeping it simpler. When STP was invented (late 1980s), bridges were pretty under-powered. Simplicity was a good thing.
Also, do you remember how slow workstations were back then? Maybe they were already so slow that nobody would notice if the bridges took a while to reconverge. Also, upper layers take a long time to close a connection anyway. They retransmit for a while before giving up. They would still be trying to keep the connection open by the time the bridges came back. Finally, the creators of STP couldn't have foreseen that we would go from networks with a few bridges to enormous flat, switched networks. These changes came about mostly because of Layer 8 marketing and financial reasons, not technical reasons. Also, today's networks are much more essential to a business's success than they were in the 80s. So today's networks have much more redundancy. Compare the diagrams in the original STP spec with the ones in the new RSTP spec. The original diagrams show a very simple bridged network with some redundancy. The new diagrams look like they come straight out of a Cisco hierarchical campus network design guide. Note the extreme level of redundancy. Priscilla At 05:16 PM 2/28/02, nrf wrote: >Cool. That's pretty much exactly what I was looking for. > >It still leaves one of my questions unanswered though. I know this might >sound like a wise-ass question, but I assure you that I ask this with no >malice. If RSTP really does offer such an improvement over STP, then why >didn't we always have RSTP? Specifically, why didn't the inventors of the >original STP put the features of RTSP into STP? Was it just a matter of >learning from a mistake - that they thought that the long STP convergence >time was acceptable, and then later realized that it wasn't? Or that they >wanted to keep STP simple, only to find that customers really wanted these >advanced (albeit complex) features? Or were there some technical issues >with switches in the old days that might have prevented proper RTSP >implementation? Or something else? > > >""Priscilla Oppenheimer"" wrote in message >news:[EMAIL PROTECTED].; > > You can get to the standard. You don't have to be a member of IEEE. > > > > filler due to url bug filler due to url bug filler due to url bug filler > > due to url bug filler due to url bug filler due to url bug filler due to > > url bug filler due to url bug filler due to url bug > > > > Near the bottom of the following page, click on Terms and Conditions and > > agree to give away your first born child if you should break these terms > > and then you can get to many standards, inlucding IEEE 802.1w. Please let > > us know what you find out. ;-) > > > > http://standards.ieee.org/getieee802/ > > > > Priscilla > > > > > > At 04:09 PM 2/28/02, nrf wrote: > > >Does anybody know exactly how Rapid Spanning Tree works, or have a link >that > > >describes it in detail? What I'm really interested in knowing is the > > >technical details that make it better than old-school STP, and in > > >particular, if RSTP is better, then why didn't the original STP designers > > >make it like RSTP in the first place (not trying to criticize, I'm just > > >interested in the evolutionary process of protocols)? > > > > > >What I find curious is that I searched and while I found that web sites > > >freely discuss how RSTP is better (or not), or talk about which vendors >have > > >implemented it or not, I haven't found a single site that describes >exactly > > >what RSTP is doing from a technical perspective and why whatever it is >doing > > >is better than STP. Furthermore, I'm not a member of IEEE, so I guess I > > >can't access the 802.1w doc. > > ________________________ > > > > Priscilla Oppenheimer > > http://www.priscilla.com ________________________ Priscilla Oppenheimer http://www.priscilla.com Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=36888&t=36851 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
