No announcement yet.

Stacking question from a newbie...

  • Filter
  • Time
  • Show
Clear All
new posts

  • Stacking question from a newbie...

    Just to clarify, I have no formal Cisco accreditiation, am merely trying to pick it up as I go along!

    We have 2 Catalyst 3750's at the site I look after, they werent installed by myself but they were put in to be Gigabit "core" switches.

    They are, logically, named XXXXXXCOR001 and XXXXXXCOR002. They appear to be "stacked" but I'm not sure to what degree. They are NOT using the stacking cables that plug into the back. Instead ports G/027 and G/028 have RJ45 adapters in them on both switches and patch cables linking them.

    I apologise if this is NOT stacking but I'm no able to contact the person whom set them up.

    COR001 has been setup as the default gateway for our internal network - All the PC's and Servers have .254 as their default gateway. This switch also has routes on it for various network ranges that are at the other end of various VPN tunnels we have. It also has the local DNS server set for name lookups but curiously also a 2nd name server which is located at the other end of one of the VPN's. This switch also has its default gateway set as which is the core firewall.

    COR002 has been setup as - - and has NOT got any routing tables on it at all and also has a default gateway on it set as

    If I look at the stack settings for both it shows them both as a master and I'm not able to change the settings.

    Also, its worth noting, my predecessor has layed things out so that each server(which all have dual NIC cards) has one NIC going to COR001 and the other going to COR002, presumably for resilience.

    My question(s) are :

    1. Should COR001 be using a name server located on the other end of a VPN tunnel or is this irrelevant in reality.

    2. Based on my description, are these switches "stacked"? And what happens in the event of COR001 failing?

    3. Even if they are stacked, if the power physically failed to COR001 or it burnt out for instance, with COR002 not having any routing tables, would things just fall over? Logic seems to suggest it would and that both core switches should have the same routing tables and default gateways and then in the event of a failure we could just change the DG for DHCP to .253. I'm guessing also, that clustering them as a proper pair would be the best way forward?

    I think thats it for now, I've probably confused the issue and apologise in advance for the basic level of description and knowledge on my part. I'm really keen to extend my knowledge and my employer isnt keen on putting me forward for the accreditation.

    Many thanks