Route Group vs. Route List for Failover

[Ali Amir]

Hi, I have a question regarding using of Route Group and Route List to manage fail over.

When we have two SIP trunks (Trunk1 and Trunk2) and want to implement fail over solution for them and keep a trunk (for example Trunk1) as main trunk, we could configure followings: but what is the best way or better to say is there any difference for following configurations:

- Configure two separate Route Groups for every trunk and add these Route Groups to a Route List. Trunk1 will be at the top in the Route List.

or

- Configure one Route Group and add both trunks to this Route Group and change Distribution Algorithm to Top Down. Then add this Route Group to a Route List.

Any Idea?

[Roger Kallberg]

One RG, with distribution set to Top Down.

[M02@rt37 · Accepted solution]

Hello @Ali Amir,

Both approaches can achieve desisred failover solution.

In my point of view, in terms of granular control, the main advantage of using separate Route Groups for each trunk lies in the ability to configure specific settings for each trunk individually. These settings can include priority, call limits, call restrictions, codec preferences, and other parameters that may be specific to a particular trunk.

On the other hand, using a single Route Group with Top Down distribution limits the ability to configure different settings for each trunk. The trunks within the Route Group will have the same attributes and priorities, and the failover behavior will be based on the order in which they are listed.

[Ibrahim AlQumairi]

RG is normally for load balancing, and RL is for failover

Create two RG for every trunk and add them to one RL 

or you can add the two trunks to one RG and change the algorithm to Top Down, but I prefer on RL

[Roger Kallberg]

Only an RG with the distribution algorithm set to Round-Robin would do load sharing, I would not go as far as call it load balancing. In my experience it's faster to do the fail-over when using one RG, that's why that was my recommendation.