FinOps Unpacked: Driving Cloud Cost Efficiency Through Collaboration

Azure Advisor offers actionable recommendations to help optimize Azure resources for reliability, security, operational excellence, performance, reduce costs, and increase the efficiency of cloud investments.

Enables the reservation of compute capacity in an Azure region or an Availability Zone for any duration of time without any commitment

Buy unused compute capacity at significant cost savings. Azure Spot Virtual Machines are great for workloads that can handle interruptions, such as batch processing jobs, dev/test environments, large compute workloads, and more.

Help save money by committing to one-year or three-year plans for multiple products. Save on Azure database services by pre-committing to fully managed services.

Azure software plans for SUSE and RedHat are reservations that apply to VMs. The software plan discount applies to the software usage of deployed virtual machines that match the reservation.

Estimates the price, savings, and eligible instances in Azure based on the organization’s current Software Assurance licenses. This enables organizations to use their existing licenses, provisioning Azure services to avoid duplicate license costs (Windows VMs, SQL Server etc.)

Serverless computing allows running code without the need to provision or manage servers. This can result in significant cost savings, only paying for the resources in use. Azure offers several serverless options, including Azure Functions, Logic Apps, and Event Grid.

Estimate the cost savings one can realize by migrating workloads to Azure.

Committing an hourly spend for Azure compute resources for a one-year or three-year plan.

Includes scripts and solutions to accelerate the Azure FinOps journey

Region Selection

Azure has multiple regions in a geography. Choose Azure resources that are cheaper in the business operational geography. For e.g., storage services are generally more affordable in us-east-2 than us-east

A Workbook to identify resources that are no longer in use. The workbook provides an overview of orphaned resources, enabling teams to enhance efficiency by:

• Cost reduction
• Prevent misconfiguration

Resource Locks prevent the accidental deletion or modification of Azure resources. This can help avoid unexpected costs due to human error. By applying a Resource Lock to a resource, the team can prevent it from being deleted or modified without an explicit override.

Automate

Use automation to

• Dynamic scaling & provisioning
• Remove unused resources and free up storage periodically
• Move data to cold and achieve storage in alignment with business needs
• Auto-tag resource with the creator’s name, type, key cost parameters, and timestamp. This will help to identify misconfigured resources and waste faster
• Remove all the resources from the Sandbox subscription periodically

3^rd Party Solutions

• CloudZero - Assigns cloud expenses to specific groups, tasks, or divisions, and uses AI algorithms to identify opportunities for cost optimization
• Finout - Tracks and identifies cost spending patterns or anomalies, and allows users to set up anomaly alerts
• FinOps hubs - Extends Cost Management to provide a scalable platform for advanced data reporting and analytics

Significantly reduce the costs of ongoing dev/test workloads with discounted rates on Azure services by using the Dev/Test environment instead of production for lower environments.

• Stop/scale down lower environments like Dev, Sandbox, QA, and UAT resources during non-operating hours
• Scale up/down as per load, on a time basis, or a mix of both

Remove costly services not required for lower environments, such as Defender for Storage (file scans), DDoS, Defender, etc.

• Do not permit developers to create resources. Implement an approval process wherein a centralized team decides on the optimum use of resources by reusing and right-sizing the resources
• Share with common network resources like Gateway, Firewall, VNet, VPN Gateway, etc.

Adopt Multi-Cloud Strategy

Evaluate and apply a multi-cloud based on where it is cheaper without affecting the business goals.

Use Artificial Intelligence

Use Custom, freely available, or commercial, ready-to-use AI/ML solutions for the following. Add their output to the dashboard.

• Workload Optimization
• Anomaly Management
• Monitoring and Alerting
• Pattern detection
• Cleanup
• Cleanup
• Recommendations

We can train ML models on historical spending data, including multiple Azure accounts. This diverse dataset allows the model to train complex relationships and produce more reliable predictions.

Optimize Networking Costs

Resource sizing

• Right-size network interface cards (NICs) to match application needs.
• Monitor bandwidth usage and adjust accordingly to avoid over-provisioning.

Traffic routing

• Leverage Azure Virtual WAN for efficient routing across multiple regions, mainly for large-scale networks.
• Use Azure VPN Gateway for site-to-site connections when appropriate.
• Implement network policies to prioritize certain traffic types to lower-cost routes.

Reserved instances

Purchase reserved instances for predictable network traffic to get significant cost discounts.

Cost management tools

• Use Azure Cost Management to identify areas of high network spending and track trends.
• Analyze network traffic patterns to identify potential cost optimization opportunities.

Data transfer optimization

• Minimize unnecessary data transfers between regions by storing data locally wherever possible.
• Compress data before transferring to reduce bandwidth usage.

Cost-effective options

• Utilize Azure peering for cost-effective data transfer within the same region. Consider spot instances for flexible workloads where price fluctuation is acceptable.

Resource-specific optimization

The following are some additional resource-specific recommendations.

• If VMs are not required regularly, then create an image and spin out a new VM from the image. Remove it after using it. Automation of this whole cycle
• Optimize logging in Application Insights and Firewall. Keep logs only for a specific period, as per business requirements
• The major cost in ADF is the choice of integration runtime. Use either the default or self-hosted integration runtime to save costs
• Use Linux VMs instead of Windows
• FTP services in Storage Accounts are relatively high. Use a single SFTP endpoint and use RBAC to control who can use it

Build a dashboard to monitor and raise alerts for preemptive care.

• Apply a suite of tools to monitor, allocate, and optimize the cost of Azure workloads. You can also use a third-party dashboarding solution and Power BI.
• Use resource management Groups for better management.

The objective is to have Comprehensive Cost visibility and alerting as a centralized source of truth in a single pane view that users can drill down to the details.