Azure K8s & MySQL Optimization

I have a Sylius / Sulu PHP backend and a Nuxt.js frontend running in Azure Kubernetes Service. All traffic ultimately reaches an Azure Database for MySQL, and under a modest load of 100–200 parallel requests each API call takes 3-4 seconds. The profiler shows the slowdown is almost entirely “slow query execution”. I have done only partial indexing and query tuning, so there is still plenty of low-hanging fruit.

What I need now is a seasoned Azure-K8s engineer who also understands MySQL performance to step in, look at the cluster and database together, and eliminate the bottlenecks. The engagement is pragmatic: measure, diagnose, fix, prove.

Key points you’ll touch:

• Audit current AKS setup (node pools, pod resources, ingress, liveness/readiness probes, connection pooling, autoscaling).
• Review Helm charts / manifests for Sylius, Sulu and Nuxt containers and tighten resource limits, env-vars and secret management.
• Trace queries on Azure Database for MySQL, add or adjust indexes, and recommend schema or query rewrites where needed.
• Tune MySQL parameters (InnoDB buffer, thread pool, query cache policies) for the workload profile.
• Implement or adjust horizontal pod autoscaler and, if necessary, a read-replica strategy to keep p95 latency under one second.
• Provide a short write-up of changes, before-and-after metrics, plus kubectl / Azure CLI or Terraform snippets so I can reproduce the environment.

I will grant you temporary access to the AKS cluster, the Azure Database for MySQL portal and our staging repository. Once your work consistently brings response times below one second under the same 100–200 request load, and the deployment remains reproducible, the task is complete.

If you are comfortable juggling AKS, Helm, MySQL EXPLAIN plans and Azure monitoring tools in one sitting, let’s get started.