Reading Time: 5 minutes

MuleSoft is delighted to announce the launch of Horizontal Autoscaling for Mule applications deployed on CloudHub 2.0 and Runtime Fabric (RTF). Managing the load placed on an application stack is one of the central challenges of scaling IT operations.

Establishing a baseline for “normal service” is one thing, but as IT organizations begin to scale, you need to account for varying demand patterns that can lead to an inordinate strain on your systems. 

Teams that have to deal with these circumstances are often shouldered with strictly monitoring application performance and having to manually scale infrastructure to accommodate traffic surges. Disruptions, slow responsiveness, and increased costs due to over-provisioning are all a risk in this scenario.

MuleSoft intends to address such challenges with the introduction of Horizontal Autoscaling.

Single-click infrastructure scaling

With the click of a button, CloudHub 2.0 and Runtime Fabric customers can now choose to automatically scale their applications to handle fluctuating demand and traffic spikes, while optimizing their Anypoint Platform usage. MuleSoft has leveraged native Kubernetes’ Horizontal Pod Autoscaler (HPA) to dynamically deploy more or fewer application replicas based on CPU utilization, ensuring seamless responsiveness to changing transaction volumes.

Runtime manager enabling horizontal autoscaling

Gone is the need to manually track load surges and redeploy applications. By setting a minimum and maximum replica range, MuleSoft automatically scales the application, reducing the operational burden and eliminating the need for custom scaling policies. 

What is horizontal autoscaling? 

Horizontal autoscaling ensures low latency scaling and automated responsiveness. MuleSoft can spin up additional replicas within minutes, allowing applications to handle traffic surges without disruptions. Scale-down operations are also carefully managed to prevent traffic loss and disruption.

Rather than upfront overprovisioning replicas, customers can leverage this feature to optimize their Anypoint Flow usage when their application is in steady state, while incurring additional usage only when their application sees an increase in demand. 

Customers can track the flows, messages and throughput usage incurred by applications enabled with autoscaling through the Usage Reports dashboard. The feature also offers increased flexibility by decoupling application scaling from infrastructure limitations, i.e. it enables customers to adopt demand based scaling without being constrained by limited infrastructure. 

Lower disruptions, greater flexibility

We take our position as a leader in the iPaaS space very seriously at MuleSoft. Our goal is to offer a market-leading degree of deployment flexibility. This extends to the control we afford our customers who choose to deploy their integrations across a variety of deployment architectures.  

As application stacks continue to scale as more businesses become application-driven, dynamism is the name of the game. Our customers need the ability to scale based on demand and focus on the value derived from their integrations rather than manually managing infrastructure.

This launch reflects that reality by providing an optimized approach to handling surges while also delivering value during scaled down, low traffic periods.