Azure FinOps Review

Cost Optimization Actions and Findings for PageGroup’s Azure Cloud Environment

Background: i conducted a cost optimization audit to identify potential savings in PageGroup's Azure cloud environment. The audit focused on various Azure services within the BI & Analytics subscription, aiming to reduce operational costs and improve efficiency.

Key Actions and Findings:

Right-Sizing Virtual Machines:

Objective: Adjust the size of virtual machines (VMs) to match workload requirements.

Findings: Significant savings can be achieved by resizing VMs. For example, resizing the virtual machine from Standard DS2_v2 to Standard B2ms could save £1,046.11 per year.

Recommendation: Implement right-sizing across all VMs to optimize costs.

Virtual Machine Reservations:

Objective: Utilize Azure reservations to commit to one-year or three-year plans for cost savings.

Findings: Purchasing reserved instances for select virtual machine could yield potential savings of £75,616 per year.

Recommendation: Purchase reserved instances for consistently utilized VMs.

SQL Database Optimization:

Objective: Optimize SQL database costs by adjusting service tiers and purchasing models.

Findings: Switching from the DTU model to the vCore model and applying reserved instances can save up to 97% per year. For example, changing a database could reduce costs from £1,731.22 to £54.41 per month.

Recommendation: Convert SQL databases to the vCore model and purchase SQL vCore reserves.

Azure Databricks Optimization:

Objective: Optimize the cost of Azure Databricks services.

Findings: Right-sizing VMs supporting Databricks and purchasing reserved Databricks Units (DBUs) can save up to 26% per year.

Recommendation: Purchase one-year reserves of 25,000 DBUs to reduce Databricks service costs.

Overall Recommendations:

Continuously monitor workloads to ensure resources are right-sized.

Implement reservations and hybrid benefits to maximize savings.

Regularly review and adjust service tiers and purchasing models to align with workload requirements.

By following the recommendations I provided, PageGroup can achieve substantial cost savings and improve the efficiency of their Azure cloud environment.