Implementing a Kubernetes Mule Platform on Azure


We improved core allocation, scalability and application health for a leading online fashion retailer to successfully address infrastructure challenges in managing Mulesoft applications by implementing a containerised Mulesoft platform in Kubernetes on Azure.


  • Cost savings of 50% on infrastructure by elastically scaling cores during normal operations.
  • Reduced operational cost by 26% by automating manual operations and support
  • Increased peak event availability (Black Friday) by 24%.
  • Increased availability from 83% to 99.99%, with:
    • RCA improvement
    • Decreased MTTR
    • 35% time savings per incident.
  • Reduced application startup time by 75% due to guaranteed vCore allocation.
  • Optimised cores allocation by 20%, improving stability and performance.


By implementing these solutions, the online retailer significantly enhanced the efficiency, scalability, and reliability of their Mule platform, ultimately delivering improved fashion experiences to customers.

  • VMs and Deployment Pipelines: The retailer faced difficulties with multiple Mule runtime versions and lacked rolling update capabilities due to VM dependency.
  • Lack of Application Isolation: Noisy neighbour issues arose due to the absence of application isolation.
  • Maintenance Burden:  Inefficient rolling updates and Mule runtime restarts burdened maintenance efforts.
  • Limited Core Allocation: Limited control over core allocation led to inefficient CPU core usage.
  • Application Health: Applications didn’t automatically restart, causing potential downtime and the need for complex health checks.
  • Elastic Scalability: Manual provisioning of VMs for additional instances hindered scalability.

The containerised Mule platform in Kubernetes on Azure addressed these issues by providing real-time monitoring, core allocation control, automated application health, elastic scalability, application isolation, and streamlined maintenance.

Third Republic