1. Home
  2. Cisco Systems
  3. preserve remote vlan with Q in Q

preserve remote vlan with Q in Q

One of our remote sites needs to move their servers to our datacenter. We need to do this while maintaining the server vlan of the remote site so the server's addresses do not change when they are moved here.

Currently the remote site is connected over a point to point fiber link. There is a 4506 on the remote site end, with a gig link configured with a 30 bit address (i.e. 10.0.1.1 255.255.255.252).

The other end connects to an intermediate site of ours into a 6509 with a corresponding address defind on the port (i.e. 10.0.1.2

255.255.255.252). The 6509 in turn is connected to our datacenter 6513 over a trunk port (connected by fiber).

We've since had another fiber line run between the remote site and our datacenter where the 6513 is housed. We can now eliminate the intermediary link from the 4506 to the 6509 and connect the remote site's 4506 directly into our 6513. How do we do this in a way that preserves the remote sites vlans (and therefore their address ranges) so that we can move their servers over here with their addresses intact? We do not want to duplicate the remote sites vlans, or have any of their vlans conflict with ours.

I've been looking at Q in Q vlan stacking but I'm not quite sure this is the right solution. I've also looked at a variation of Q in Q that uses that subinterfaces for vlan mapping from customer to service provider vlans, but I'm not sure this is the right solution either.

What is the best way to do this? Please ask questions if more information is needed.

Reply to
willsmith1701
Loading thread data ...

the easiest way is to trunk the server vlan thru to the new server location. local user vlans can stay at the original site. this doesnt need Q in Q - it only needs you to trunk that vlan thru the intermediate links.

Note this means your GigE WAN links now use vlan tags - you might need to reconfigure them to make that happen, and you need to be careful that other vlans do not "bleed" across the links.

so long as there is a layer 2 path between the servers and the routers on the subnet they originally were connected to, then everything is going to work, although you might find some L3 traffic flows crossing WAN links 2 or more times.

However - if you already use the server vlan number in any other site you will find the 2 or more vlans merging - whether that is a problem depends on the network design.

Q in Q just means that you can avoid a vlan number clash at the intermediate site.

the way to solve this problem properly is to use a naming scheme that makes the IP address "changeable" - any system which depepnds on you not altering addresses is going to bite at some point.

did the people that told you that you cannot change the server IP addresses know what they were talking about?

Reply to
stephen

One way is virtualization via a routed link and their own VRF. Depends on network design otherwise, and what the need access to. But VRF rocks !

HTH Martin

Reply to
Martin Bilgrav

Inventory the VLAN numbers used at the remote site to ensure that they do not conflict with any of the datacenter VLANs.

Bring up the new link between the 4506 and the 6513 as a .1q trunk and allow only the VLANs required.

Reply to
Merv

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.