Eric, We went through the same dilemma. We are upgrading from fully redundant 5500's to 6500's (Dual Sup2, MSFC2, PFC2). We decided against HSRP because it required us to keep the configurations synchronized. Granted, the auto sync would take care of IP and IPX, but we also have Appletalk which would require extra effort to maintain (disable sync, make the change to each MSFC, enable sync).
The downside to SRM is the failover time. Although we haven't actually tested the failover time under load, Cisco says it could take up to 120 seconds while all the tables are repopulated on the redundant Sup/MSFC. Existing flows continue to run while this happens, just new flows can't be established. We decided that was acceptable. As for packet loss, I don't think the 6500 would have problems unless you're passing 10's of Gig's per second on a non-fabric enabled switch. We are supporting upwards of 25,000 Ethernet ports with over 100 VLANs. We have not had any problems with EIGRP overloading the 5500's or 6500. We decided to use passive on all VLANs except those that actually need it. This is more for security than anything else. I'm curious... is the 65007 a James Bond switch? :-) Ken >>> "Eric W" 11/28/02 04:47PM >>> Dear Cisco Fans and Professionals, I need some friendly advice. There are different opinions about Cat6500(High availibility with Single Router Mode) and (High availibility with Dual MSFC Redundancy) Imagine you had 3 Cat65007s with Dual MSFC1 and Dual Supervisor1A. That is 6 MSFC's and 6 Supervisor engines. You have over 1500 users to support with 30VLANs that need interconnecting routing via EIGRP. Some user applications are sensitive to packet loss. You have to provide minimium downtime in the case of MSFC failer or Supervisor failer. Question 1. Which would you implement (High availibility with Single router mode) or (High availibility with Dual MSFC Redundancy) and why? Question 2. Is was brought to my attention that running dual MSFC redundancy with a high number of VLANS would cause the EIGRP process to run very high. As you know EIGRP is a very noisy protocol. Query storms.. Implementing EIGRP passive interface on all VLAN interfaces except the management VLAN would help the EIGRP process to run low. Are query storms an issue to worry about even after the passive interface is issued on client/user VLANS? Regards, Eric Washington Network Engineer CCNP Thanks in advance for your input Cisco Professionals _________________________________________________________________ MSN Hotmail http://www.hotmail.com/ BEGIN:VCARD VERSION:2.1 X-GWTYPE:USER FN:Ken Diliberto EMAIL;WORK;PREF;NGW:[EMAIL PROTECTED] N:Diliberto;Ken END:VCARD Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=58276&t=58276 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]