Redundant LTM3600 pair on different subnets

The only way to set it up is via the GUI.

 

 

Here is the following criteria for setting up HA (Basic setup)

 

1. You need to connect the 2 LTMs via Serial Cable (Special one provided by F5.).

 

2. Create a VLAN and create a self-address on each

 

Example:

 

BIGIP1

 

- VLAN Internal

 

- Self Address: 192.168.1.2

 

- Floating Address: 192.168.1.1

 

- VLAN External

 

- Self Address: 192.168.2.2

 

- Floating Address: 192.168.2.1

 

 

BIGIP1

 

- VLAN Internal

 

- Self Address: 192.168.1.3

 

- Floating Address: 192.168.1.1

 

- VLAN External

 

- Self Address: 192.168.2.3

 

- Floating Address: 192.168.2.1

 

 

 

 

Hope this helps.

 

CB

 

How would that work? Both LTMs are on different subnets (network/locations)? They cannot be connected through the serial cable.

I may be missing something here but I'm not sure you can do this. The way BigIP deals with failovers is by issuing a GARP onto the wire, which will only be useful if the units are adjacent to each other L2-wise. What is your specific goal?

 

 

-Matt

I have only 2 LTMs on east coast. I need to build a disaster recovery (failover) site and take 1 LTM to west coast. I don't have the budget to buy new equipments but to use only existing ones. How can i accomplish an automatic failover if LTM on east coast is DOWN or unreachable?

 

 

Here is what i have been fed but i don't know how:

 

1 LTM in east

 

1 LTM in west

 

 

Use BGP/IGP on upstream routers to load balance the traffic and set priority so that when both are UP, the preferred path is east.

 

Can someone elaborate and explain different component required in this?

 

You really need a GTM at both sites to properly accomplish site-to-site failover.

 

 

There's no way to set up an active/standby LTM pair with different IP addresses for the shared components. You could potentially do this with just using each LTM as a standalone (I wouldn't recommend active/active over that distance either) and then BGP as you described, but I don't see a way to do this with the LTM's set up as a redundant pair.

 

 

Denny

If i can't setup Active/Standby LTM pain with different ip addresses then what is the significance of having Network Failover in the "high availability > redundancy" section ?

Network fail over can be used for large layer 2 spanned networks (between two sites, for example), for environments where the distance between the BigIPs is too far for the serial fail over cable (e.g. a data center with "a" and "b" cabinets physically separated), or for scenarios where you want two fail over mechanisms. All of this is predicated upon the same network block being configured on the systems.

 

 

I agree with Denny: GTM is the best solution here.

 

 

Regards,

 

Matt

While I think it's theoritically possible to do an active/standby on the same network accross 2 geographic sites, either through a switch with spanned VLAN accross a GRE or a IPSEC Tunnel that carries tagged VLAN. It's too elaborate a setup to support this mechanism. I am not even sure F5 would be able to support something it was never designed to do. Thus I agree with Denny and Matt that a GTM or a GTM service can provide you a solution.

 

 

CB

I have data center A in 156.118.x.1 subnet

 

and data center B in 156.118.y.1 subnet.

 

 

 

Please explain layer 2 spanned network ?

 

 

I agree GTM is the best sol but i need to make it work in the interim.

Posted By navgup on 04/13/2009 11:39 AM

 

Please explain layer 2 spanned network ?

 

 

 

Both data centers would need to be on the same Layer 2 subnet, ie, both on 156.118.x.0. Making that happen is probably a lot more expensive than a couple of GTM's.

 

 

It's not just a problem with the network failover. It's also because you could not share the same virtual server (or any other shared address - SNAT, gateway IP, etc.) across anything but a Layer 2 subnet.

 

 

Denny