interesting question - a thought or two in line
""Azhar Teza"" wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > I have a customer who has (1) 6509 backbone switches which is also acting > as a root bridge. (7) 3524 are connect back to 6509 via fiber each as a > seperate unit. At this point, if 6509 fails whole network will go down. I > suggested to have an additional switch run in standby mode as a backup > backbone switch. Customer doesn't want to spend around 40,000 to buy a 2nd > 6509 switch. Can the backup switch be another Catalyst say 4000 or does it > have to be the exact same model. I know the performance will be downgraded > since 4000 series don't have the same switching backplane as opposed to > 6509, but still it should take over as a root bridge incase 6509 goes > down. In this configuration, the only thing they will have to do is to > move their servers to the 4006 switch until the 6509 comes back online. All > I need to make sure that the both 6509 and 4006 switch have the same > configuration. Is there anything I am missing, please shed some lights > guys. Teza CL: what kind of failure are you protecting against? I understand that everything homes back into the 6509. I'm just wondering about the connectivy requirements. CL: servers connect to what? CL: other essential services connect to what? CL: to protect against single points of failure, you will need to consider the following: 1) redundant 6509 2) dual homing of servers 3) dual homing of all other switches, to each of the two 6509's CL: if that is too pricy, another thought might be this: 1) segment all servers onto their own switch - a 3550-24 or 48 or 12 - whichever is appropirate. 2) get a 3550-12G, make it the root bridge, and dual home the server switch to both the 3550-12G and the 6509. Now, if the 6509 fails, other segments will have connectivity to the servers. If the 3550-12G fails, the 6509 allows server connectivity to those connected to it. 3) dual home all your closet switches to the 3550-12G and to the 6509. CL: under this scneario, you still have a single point of failure in the box that the servers are connected to. of course, the world is a single point of failure ;-> CL: just a thought. maybe not the best idea, but certainly effective and inexpensive. > > ------------------------------------------------ > Changed your e-mail? Keep your contacts! Use this free e-mail change of > address service from Return Path. Register now! Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=54615&t=54614 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
