Dave,
You are correct. Same setup I have working on many 6509's and 4006's today. 4006's dual ran to each 6509. Each 6509 is my VTP Server. I can take one completely down for maintenance or whatever and it will never miss a packet. Under than the first 50 seconds for spanning tree to go from a listening---learning-forwarding state... When I said "Vlans would basically go down" What I meant is with you VTP server down no vlan routing can take place because the server announcing that vlan is down. If you were to have 2 clients on different vlans on a 4006 and you take down your VTP server. Client A in VLAN A could not get to Client B in Vlan B. All local vlan traffic would work on the 4006 because it is locally switched but it cannot route vlans. Ronnie ""The Road Goes Ever On"" wrote in message news:[EMAIL PROTECTED] > ""Ronnie Higginbotham"" wrote in message > news:[EMAIL PROTECTED] > > Once you take the VTP server down all vlans will basically go down. Which > > will bring down the network with only one VTP server. They will sit there > > and do nothing until the vtp server comes back online. > > Is this right? I got to wondering about this, but I don't have enough > equipment to simulate a real network to test end to end. > > What I did do was take my two switches, set one as the VTP server, the other > as the client, create some vlans on the server, assigning those vlans to > ports on the client switch, shut down the trunk ports between the two > switches, and watch. > > I'm not sure how long I need to wait, but what I am seeing ( or not seeing ) > confirms my suspicion - that the assigned ports on the client switch remain > functional. > > My trunks are down on the client switch, but that is expected in this setup. > > What I am unable to test is whether or not the trunks, over which vlan > information is propagated, will still function in the scenario described. > > I suppose you are correct when you say that "all vlans will basically go > down". Certainly, in my simulation, no one on a local switch could get to > the core. What I am wondering is in a dual core switch situation as > described, if servers and services were dual homed to the redundant > switches, if all would remain well. Certainly no new vlans could be added, > but I'm wondering if all else would remain functioning. > > some info on the switch > > Switch_48#sh vtp stat > VTP Version : 2 > Configuration Revision : 2 > Maximum VLANs supported locally : 1005 > Number of existing VLANs : 12 > VTP Operating Mode : Client > > Switch_48#sh vlan > > VLAN Name Status Ports > ---- -------------------------------- --------- -------------------------- -- > --- > 1 default active Fa0/4, Fa0/5, Fa0/6, Fa0/7 > Fa0/11, Fa0/12, Fa0/13, > Fa0/14 > edited for brevety > > 10 VLAN0010 active Fa0/10 > 20 VLAN0020 active Fa0/20 > 30 VLAN0030 active Fa0/8, Fa0/30 > 40 VLAN0040 active Fa0/40 > 100 VLAN0100 active Fa0/1 > 200 VLAN0200 active Fa0/2 > 300 VLAN0300 active Fa0/3 > 1002 fddi-default active > 1003 trcrf-default active > --M > > Switch_48#sh int trunk > > Switch_48# ( demonstrating that the trunk is down ) > > > As I said, with two switches, I can't test a real world. However, I suspect > that if the network is properly design, with dual homing of servers and > services, and dual home of the access closets to the core switch, that > things might just be fine. > > Can anyone provide info in a more realistic simulation? > > > >It would be a good > > idea to have two VTP servers so you can take them down one at a time and > > upgrade. > > always a good idea. > > > I must assume the 2 6509's or for you core routing and the 4006's > > are your closet access switches. > > > > If the other standby 6509 is a client in the same VTP domain as you server > > which I don't see why it would not be, they already that the same VTP > > revision all you have to do is promote the client 6509 to server. > > > > set vtp domain (Domain Name) mode server > > > > Then you have two servers. You must have the 4006's dual ran to both > 6509's > > to work... > > > > Ronnie > > > > ""Dave C."" wrote in message > > news:[EMAIL PROTECTED] > > > I have a question regarding VTP. > > > > > > I am working with a network configuration that has dual 6509's as core > > > switches. One is the VTP domain server, the other is a VTP domain > client. > > > All other 4000 switches are defined as client also. > > > > > > I need to take both 6509's down to upgrade the Cat-OS on them, however I > > am > > > concerned about what happens to the other switches when the VTP domain > > > server goes away while the other 6509 is in client mode. > > > > > > I know that I can assign multiple VTP servers to a domain. My original > > > thought is to first make sure that the client 6509 is running the same > VTP > > > revision # as the server 6509. Then I should be able to change the > client > > > 6509 to VTP server mode. Then in theory either server can take over VTP > > > responsibilites, because one will be able to act as the other if the > other > > > one goes away. > > > > > > What I was also wondering, do all VLAN's go disabled if I were to only > > have > > > a single VTP server and shut it down when the rest of the switches are > > > clients? If so, would this happen right away, or after 5 minutes? If I > > take > > > the 6509 server down, it will not be able to tell anyone that a topology > > > change has occurred, but will the clients be looking for the VTP 5 > minute > > > updates? > > > > > > If anyone could help clarify this, I would appreciate it. > > > > > > Thanks. > > > > > > Dave C. Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=70964&t=70942 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
