Question: Redundancy with Local Route Groups

[tschafferx]

Hello Cisco community,

I was wondering, how a redundancy could be set up with the use of LRGs. How would I configure the following scenario with LRGs:

1 Route Pattern -- 1 Route List containing 2 individual RGs --- RG 1 IP-WAN - RG 2 PSTN (Failover) -- Devices

With LRGs I can only assign 1 RG at the Device Pool level. What if the devices in that RG go down and I want to provide a backup path over the PSTN.

Thank you in advance.

Any help is appreciated.

[R0g22 · Accepted solution]

You would leverage the route list for redundancy in this case -

RL -
Standard Local Route Group (1st Preference)
PSTN-RG (2nd Preference)

[tschafferx]

Thank you for the reply. Okay, then it's important to set the distribution algorithm to Top Down. I guess that with LRGs load sharing via a circular approach would not be feasible. Correct?

[R0g22]

There is no algorithm in a route list that you can configure on CUCM. Distribution algo is there for RG.

[tschafferx]

Thank you for the quick reply.

I am aware of that. But there is for Route Groups and if I wanted to achieve some sort of load balancing with LRGs I would have to set the distribution algorithm (within the respective Route list) to circular, which in turn would invalidate the design for redundancy.

[R0g22]

Not sure if I understand you but per your requirement you were looking for a way to provide redundancy when using LRG which I highlighted that can be done within the RL itself. Now for example the PSTN-RG that I have mentioned above have multiple gateways/trunks etc, yes the distribution algorithm will come into picture which will only come into effect if the SLRG is not able to route the call/fails.

With the RL redundancy you aren't invalidating anything. Again, distribution is not within the RL but its something that is within the RG.

[tschafferx]

The first part is clear. You mentioned that redundancy can be achieved by placing several devices into the LRG. That's understood. What I did now is to compare it to the old approach with RL -- RG -- Devices where I could have several RGs under 1 RL with different purposes. So let's say that RG 1 is assigned to 2 GWs that have a circular distribution algorithm to achieve load balancing over the two GWs. RG 2 is assigned to a SIP-Trunk that I want to use, in case the 2 Gas go down. CUCM would see that within the RL there are two RGs selected by priority. 1st RG goes down it would automatically switch to the 2nd RG. 

Now I thought about a deployment to copy that behavior with LRGs. What I have seen is that with LRGs you can only specify 1 Route Group with various devices but not opposed to the 'traditional' approach several RGs inside a single RL.

I hope that makes it more clear. :)

[R0g22]

Understood and your are correct but there with a close mimic to what you are trying to achieve would be what I have explained in my initial post.

[tschafferx]

Okay, got it. :) Thank you! I was totally focused on LRG and thought you mean the 1 route list within the LRG scenario... That's what confused me, I couldn't see where the two RGs would be placed within the LRG configuration. The reason for posting the question was to find out, if there's maybe a way to achieve the described that I haven't been aware of yet.