On-Premise Data Gateway - Alternative Methods for Load Balancing

[PC_PBI]

I am designing On-Premise Data Gateway architecture. 

For failover and load balancing, I will be creating a Gateway cluster with 2 Gateway servers and configuring it for load balancing on the service. 

Is this the only available method to implement Load Balancing for On-Premise Data Gateway? Are there alternative designs to consider? 

Thank you

[Greg_Deckler]

@PC_PBI Well, you could go old school and put your gateways in different clusters on different servers and then use a load balancer like an F5 or something but not sure why you would complicate your life like that.

[PC_PBI]

Thanks for the info

This will help the conversation. 

[blopez11]

For out of the box load balancing, there is random, distribute, or you can load balance based on cpu and memory throttling

https://docs.microsoft.com/en-us/data-integration/gateway/service-gateway-high-availability-clusters...

I am only using the distribute option, so not sure how good the load balancing works based on cpu and memory