Multi-Cloud MongoDB Production Scaling

Context

Production MongoDB cluster spanning AWS and Azure, supporting a workload consolidation effort in which previously isolated environments were unified onto shared infrastructure.

Problem

Aggregated load on the cluster after consolidation pushed CPU, memory and disk IOPS toward capacity headroom. The cluster needed to be resized and re-provisioned across both clouds without disrupting production traffic.

My role

Operator on the scaling plan: validating instance sizing, coordinating execution windows across both clouds, and confirming health between steps.

Technical actions

1. [01] Resized 9 production MongoDB nodes across AWS availability zones and Azure regions.
2. [02] Increased disk IOPS provisioning across the fleet to meet the post-consolidation working set.
3. [03] Coordinated rolling instance replacements to keep the cluster online and the replica set healthy at every step.
4. [04] Validated cluster state, replication lag and read/write availability between phases.
5. [05] Documented the executed plan and the rollback path for each cloud-specific step.

Operational impact

Cluster sized to absorb the consolidated workload. Cross-cloud capacity planning executed without measurable production disruption.

Evidence

• [✓] 9 production MongoDB nodes resized across AWS availability zones and Azure regions.
• [✓] Disk IOPS provisioning increased fleet-wide to match the post-consolidation working set.
• [✓] Rolling instance replacements kept the replica set healthy at every step.
• [✓] Cluster state, replication lag and read/write availability validated between phases.
• [✓] Cloud-specific rollback path documented for each step.

What this demonstrates

• Production database scaling with replica-set discipline.
• Multi-cloud operational coordination (AWS + Azure).
• Capacity planning around consolidated workloads instead of isolated baselines.
• Operational responsibility on a high-stakes change.