Link redundancy over two point to point links

[Adam M]

Hello Forum, 

We have two sites:

1. Data Centre

2. Head office

We have two point to point WAN services (point to point) that are connected to the Data centre and to the Head Office:

1. 10gbps dark fibre (primary)

2. 1gbps ethernet (secondary)

In the data centre we have two Cisco 3750X switches that are stacked. Also, in the head office we have two 3750X switches that are stacked.

Each WAN service is connected to the following ports of each switch.

DC-Switch1-1/1/1 (10gbps)

DC-Switch2-2/0/48 (1gbps)

HO-Switch1-1/1/1 (10gbps)

HO-Switch2-2/0/48 (1gbps)

I would like to have the 10gbps link as my primary, and if it goes down for whatever reason, the 1gbps link will become active until the primary is available again. Once the primary is available it should fail back over to it.

Can someone pleas help me understand how I should be configuring this?

[Joseph W. Doherty]

How are you using the links logically?  I.e. L2, or L3?  What's the logical topology?

[awiadministration]

I'd like to know what is best practice really. The switches in the head office are access switches, and the switches in the data centre connect to all our server and network resources.

What would be the best topology?

Thanks

[Joseph W. Doherty]

Disclaimer

The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages wha2tsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.

Posting

Not sure there's really a "best practice", but there's a preferences for L3 over L2, when possible.

With a L3 topology, rerouting traffic if link drops or comes on-line, is generally done by a dynamic routing protocol.  "Tuned" OSPF or EIGRP can re-converge quickly, perhaps under a second.

[Adam M]

Thanks for the response and guidance to use a layer 3 routing protocol such as OSPF for link redundancy.

The switches in the head office are 'access'; user facing. So I'm guessing I would be making any user vlan (subnet) connected to that HO switch routable (layer 3) using inter-vlan routing. Is that correct?

The users would be in a /24 network (subnet) and I would make it's gateway the HO switch, then via inter-vlan routing the traffic would connect via the /29 OSPF subnets to the DC switches. Is that correct?

What would the OSPF config look like?

Attached is a diagram.

Thanks so much

[Joseph W. Doherty]

Disclaimer

The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages wha2tsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.

Posting

Well you could do all as you describe, but the links between the HO and DC could be p2p routed links, either /30s or /31s.

If you only have need for one subnet in the HO, there would be no inter-VLAN routing, although there would be routing between the HO and DC stacks.

I would suggest you avoid using VLAN 1 for a user VLAN.

You've mixed 10.x.x.x and 196.168.x.x subnets, which will work, but you could use just 10.x.x.x address space.  If you did, you would have the option to have all interfaces that have a 10.x.x.x/16 IP into the OSPF topology.

A basic OSPF config might be something like:

DC stack

int lo0

ip address 10.0.0.1 255.255.255.255

int t1/1/1

no switchport

ip address 10.1.1.1 255.255.255.254

int g2/1/1

no switchport

ip address 10.1.2.1 255.255.255.254

router ospf 10

network 10.0.0.0 0.0.255.255

HO Stack

int lo0

ip address 10.0.0.2 255.255.255.255

int t1/1/1

no switchport

ip address 10.1.1.2 255.255.255.254

int g2/1/1

no switchport

ip address 10.1.2.2 255.255.255.254

int vl10

ip address 10.2.1.1 255.255.255.0

router ospf 10

network 10.0.0.0 0.0.255.255

passive-interface vl10