I currently look after 6 x 6509 switches, 2 are dedicated core switches (which I'll call (a) and (b)), whilst the other 4 are edge switches. One of the core switches is a VTP server (a), the other 5 all vtp clients. Going through the configs, I've noticed a few areas I need some clarity over.
Firstly, I have a large number of VLANs (15 in total), each with IP addresses configured from the subnets in each vlan on both Core Switches (i.e. Vlan 10 has on (a) ip address 10.x.x.2 and on (b) ip address 10.x.x.3 with an HSRP of 10.x.x.1 on both, 20.x.x.2 (a) and 20.x.x.3 (b) and an HSRP 20.x.x.1 for vlan 20 etc). In a nutshell, each vlan taking the first 3 ip addresses of the subnet.
At present, as the second core switch is a vtp client, I am not able to configure an IP address within a new vlan I create on the second switch to remain consistent with the current config, so I'm assuming this switch must have been configured as a VTP server at some point. That said, is it necessary for HSRP to have 2 ip addresses (one on each vlan on the 2 cores) to work properly, or is this in error and why hasn't the VTP domain removed these addresses?
I'd also like to know the impact on routing should HSRP become active. At present our DHCP server configures clients to point to the DG as the HSRP address (10.x.x.1), yet when performing a trace route, it hops to the ip address of the vlan (10.x.x.2) and ignores the HSRP address.
Say for example, if that switch went offline in the current configuration, would a trace route hop to 10.x.x.3 instead of 10.x.x.2 or does the .2 stay active (broadcast) as a result of its VTP domain membership.
Please advise as I'm largely concerned about the current configuration. If you need further clarity on anything, please advise.