Skip to main content

Optimize Azure Virtual Machines

Azure Virtual Machines (VMs) are a core service in Microsoft Azure, enabling you to run a wide range of workloads in the cloud. Proper configuration and optimization of your VMs are crucial for security, performance, and cost efficiency.

In this guide, we'll explore best practices and recommendations for securing, optimizing, and managing your Azure Virtual Machines.

Refer to the Managed Disk Optimization Guide on how to optimize storage for VMs.

Optimize Your Azure Environment with Us!
Looking to enhance your Azure setup for cost efficiency, performance, reliability, or security?
Talk to an Azure expert! Email us or schedule a 30-minute consultation and let's optimize your Azure environment together!
Stay ahead with actionable insights for Azure optimization. Subscribe to updates and unlock the full potential of Azure!

Cost Optimization Recommendations​

Deallocate Stopped VMs to Save Costs​

Impact: Medium

When VMs are stopped but not deallocated, you still incur charges for allocated resources like CPU and memory. Consider deallocating VMs when not in use to avoid unnecessary costs.

Review your VM usage patterns and implement automation to deallocate unused VMs.

Optimize VM Processor Architecture​

Impact: Low

For certain workloads, switching to AMD processors instead of Intel can offer a better price-to-performance ratio. Review your workloads and test the performance of different processor architectures to determine if cost savings can be achieved.

Review VM processor architecture options to optimize costs.

Use Spot VMs for Interruptible Workloads​

Impact: Low

For non-critical workloads, consider using Azure Spot VMs. These VMs offer significant savings by using excess capacity in Azure but can be interrupted at any time, making them ideal for batch processing and dev/test environments.

Consider a mixed approach using both Standard and Spot VMs to optimize costs.

Implement Auto-Stop for Non-Critical VMs​

Impact: Low

Implement auto-shutdown for non-critical VMs to save on costs. Configure automatic shutdown based on your operating hours to ensure VMs are not running unnecessarily.

Configure VM auto-shutdown for non-critical workloads to reduce cost.

Performance Recommendations​

Choose Appropriate VM Sizes and SKUs​

Impact: High

Selecting the right VM size and SKU based on workload requirements is crucial for both performance and cost optimization. Analyze your workload's CPU, memory, and I/O requirements, and choose the best-fitting VM size.

Leverage VM Scale Sets for Auto-scaling​

Impact: High

Use Virtual Machine Scale Sets (VMSS) to automatically scale VMs based on demand. This ensures that your application can handle traffic spikes without manual intervention while optimizing resource usage during off-peak periods.

Configure auto-scaling rules based on metrics such as CPU utilization or request count.

Optimize VM Storage​

Impact: Medium

Ensure your VM storage (OS and data disks) is properly sized and optimized. Avoid storing application data on OS disks, as it can degrade performance. Use Azure Premium Storage for high-performance workloads and Azure Standard Storage for less demanding workloads.

Consider resizing your disks based on usage patterns and expected growth.

Reliability Recommendations​

Enable Azure Site Recovery (ASR)​

Impact: High

Ensure that critical VMs are protected using Azure Site Recovery (ASR). This provides disaster recovery capabilities, replicating your VMs to another region or Availability Zone.

Review all VMs for ASR protection and implement disaster recovery plans.

Use Availability Zones for High Availability​

Impact: High

Deploy your VMs in Availability Zones to ensure high availability and fault tolerance. Availability Zones are physically separate locations within an Azure region, reducing the risk of a single point of failure.

Move or redeploy VMs to Availability Zones for improved resiliency.

Implement Backup for Critical VMs​

Impact: High

Enable Azure Backup for your critical VMs. Ensure that all VMs are backed up regularly to protect against data loss and to maintain business continuity in case of failures.

Review and configure backup policies to ensure your VMs are adequately protected.

Review Virtual Machine OS Disk Size​

Impact: Low

If a VM has a large OS disk (greater than 256 GB) without any attached data disks, consider adding separate data disks. Storing data on the OS disk can cause performance degradation and complicate disaster recovery.

Ensure OS disks are used solely for the OS and system files, while application data is stored on separate data disks.

Security Recommendations​

Regularly Apply OS Patches​

Impact: High

Ensure that your Virtual Machines are regularly patched to address security vulnerabilities. A VM with outdated OS patches can expose your environment to known security risks, so it's critical to maintain up-to-date software.

Enable automatic updates for the OS and review the patch status frequently.

Protect VMs with NSGs​

Impact: High

Network Security Groups (NSGs) should be configured to restrict access to your VMs. Only allow necessary inbound and outbound traffic. Ensure that management ports like RDP (3389) and SSH (22) are tightly controlled and ideally only accessible from trusted IP addresses.

Consider using a bastion host for secure RDP/SSH access and blocking direct public access to the VMs.

Limit Public IP Address Exposure​

Impact: Medium

Avoid using public IP addresses on VMs unless absolutely necessary. If public IPs are needed, place the VM behind a Firewall, Load Balancer, or Application Gateway to restrict direct internet access and mitigate risks.

Review your VMs and consider using private IP addresses wherever possible.

Operational Excellence Recommendations​

Review Orphaned Availability Sets​

Impact: Low

Availability sets with no virtual machines are considered unused and should be reviewed for deletion to support operational excellence and reduce resource clutter. Deleting unused resources ensures a more streamlined environment.

Review orphaned Availability Sets and delete unnecessary resources to maintain a cleaner and more efficient cloud environment.

Recommendation Engine​

The Cloudconomist AI recommendation engine scans the usage metrics of your VM over the past 30 days to identify patterns and provide optimization recommendations such as:

  • High CPU Utilization - When a VM is experiencing critical CPU utilization, with usage exceeding 85% for 95% of the time over the past 30 days. Immediate action is recommended to prevent service degradation.

  • Medium CPU Utilization - When a VM's CPU utilization is between 70% and 85% for 95% of the time over the past 30 days. Consider scaling up or investigating performance if this pattern continues.

  • Low CPU Utilization - When a VM's CPU utilization is below 30% for 95% of the time over the past 30 days. The VM may be over-provisioned, and downsizing is recommended to optimize costs.

  • Network Traffic Spike - A sudden spike in network traffic has been detected. Investigate for potential security issues.

  • High OS Disk IOPS - OS disk IOPS utilization exceeds 70%. Consider scaling up or investigating performance.

  • Low OS Disk IOPS with Premium Storage - OS disk IOPS utilization is below 20% while using Premium Storage. Consider scaling down to a Standard disk to save costs. Note that the SLA for the VM will reduce from 99.9% to 99.5%.

  • High Data Disk IOPS - Data disk IOPS utilization exceeds 70%. Consider scaling up or investigating performance.

  • Low Data Disk IOPS with Premium Storage - Data disk IOPS utilization is below 20% while using Premium Storage. Consider scaling down to a Standard disk to save costs. Note that the SLA for the VM will reduce from 99.9% to 99.5%.