Add new "messageReceiveRateTarget" metric to Solace Scaler

[dennis-brinley]

Proposal

Add a new messageReceiveRateTarget metric to Solace Scaler. This metric will allow KEDA to scale workloads based on the rate of message delivery to Solace PubSub+ queues.

Use-Case

This enhancement will permit KEDA to maintain workloads at a level desired measured by the rate of message ingress to queues. Thus reducing the tendency of flapping (creation and destruction of pods) as queue backlogs are cleared.
Currently available metrics of messageCountTarget and messageSpoolUseTarget allow KEDA to scale workloads based upon existing backlogs. The new metric allows KEDA to scale workloads based upon the rate of message ingress. This feature has two benefits:

• Provides a mechanism to maintain performant applications when existing backlogs are small
• Provides a method to reduce pod "flapping," the continuous creation and destruction of pods, as KEDA maintains workload replicas in concert with observed demand in the form of queue backlogs.

Is this a feature you are interested in implementing yourself?

Yes

Anything else?

No response