January 2024 Summaries

Vantage has launched Amazon EC2 rightsizing recommendations to help users identify and address over-provisioned EC2 instances, which often result in unnecessary costs due to over-allocation of resources like vCPU and RAM. Previously, Vantage offered cost recommendations focused on generational upgrades, but the new feature provides insights specifically for rightsizing, allowing users to see potential savings and the number of instances that could be optimized. The recommendations are accessible via the Cost Recommendations page, and a new Rightsizing tab on the Active Resources screen provides details such as recommended migration sizes and reasons for rightsizing, with an option to compare pricing against current instance types. This feature is available at no additional cost to Vantage users with an AWS connection, although details may be limited for those on a free trial.

AWS Elastic File System (EFS) offers three storage classes—Standard, Infrequent Access, and the newly introduced Archive—each catering to different performance and cost needs. EFS Archive, the most economical option, is suitable for data accessed infrequently, offering up to 97% savings compared to EFS Standard. Users can transition files across these classes using lifecycle management to optimize costs based on data access frequency. EFS also provides two file system options: Regional, which offers high availability by storing files across multiple availability zones, and One Zone, which is more cost-effective but with lower availability. A case study involving a university research project demonstrates significant cost savings by strategically using these storage classes, particularly when transitioning from Standard directly to Archive. Understanding the trade-offs between availability, cost, and performance is crucial for selecting the most appropriate storage class and file system to meet specific use cases and budget requirements.

Vantage has introduced a new feature in its Cost Reports that allows users to filter and group expenses by a newly added "Charge Type" dimension, which includes categories like upfront charges, usage, amortized fees, taxes, and credits. Previously, users could adjust Cost Report settings to manage credits, taxes, refunds, and amortization, but they lacked the capacity to filter or group by specific charge types, making it challenging to associate resources with particular expenses. This enhancement enables users to distinctly analyze and separate upfront purchases from usage, facilitating more precise budget allocation and resource management. The Charge Type filter and grouping options are accessible to all Vantage customers at no additional cost, and they are supported by all providers currently available in Vantage. The feature is particularly beneficial for enterprise customers involved in significant discount programs, such as the AWS Enterprise Discount Program (EDP). Users can apply this functionality via the Vantage interface or API, making it a versatile tool for optimizing cloud costs.

In an effort to simplify Kubernetes cost monitoring, a standalone Vantage Kubernetes agent was developed to address inefficiencies and dependencies seen in prior integrations like OpenCost. This agent aligns with the Unix Philosophy of simplicity and modularity by eliminating the need for intermediary tools such as Prometheus, which previously added complexity and resource demands, especially for large-scale clusters. The new agent, inspired by metrics-server principles, effectively reduces CPU and memory usage by 95% and 97% respectively compared to earlier solutions, while providing quick setup, enhanced scalability, and precise cost data directly from the kube-apiserver. This approach not only streamlines the deployment process but also aligns cost monitoring with customers’ expectations by leveraging existing cloud cost data, thereby offering a more efficient and accurate solution for tracking Kubernetes costs across any cloud provider.

Vantage has introduced a new Tag dimension feature for Advanced Cost Report exports, enabling users to group export data by specific categories such as individual business units or projects. This enhancement offers more flexibility and control over cost data by allowing the selection of a Tag key, which adds a corresponding column with associated Tag values to the exported CSV file. Previously, grouping options were limited to attributes like Region or Service category, but the inclusion of Tags facilitates detailed analysis for organizations with multiple projects across various units. This feature is available at no extra cost and can be accessed via the Advanced export option in the Vantage console, with exports currently limited to CSV format. While only one Tag key can be selected at a time, Vantage plans to expand this functionality in the future.

MongoDB's cost optimization strategies focus on addressing common inefficiencies observed in its usage across various implementations, emphasizing the importance of selecting the appropriate pricing tier, managing data transfer costs, and minimizing data storage expenses. MongoDB offers different versions, including the free Community and the feature-rich Enterprise editions, with MongoDB Atlas providing cloud-based options across AWS, Azure, and Google Cloud. Atlas has three pricing tiers: Serverless, Shared, and Dedicated, each with unique cost structures and use cases. Serverless is ideal for low and fluctuating workloads but requires careful query optimization to prevent unexpected expenses. Data transfer and storage costs can be reduced by using local caching, paginated responses, efficient schema designs, and storage compression. Dedicated Instances can be optimized through instance size selection, cluster auto-scaling, and sharded clusters. Understanding the pricing model and leveraging built-in tools for monitoring and alerts are crucial for maintaining sustainable MongoDB expenditure.

Vantage has announced integration support for PlanetScale costs, allowing users to monitor and optimize expenses related to this database-as-a-service alongside other infrastructure costs in the Vantage console. Users can connect their PlanetScale accounts through an OAuth workflow, enabling Vantage to ingest and visualize cost data, which helps in identifying cost drivers and making informed optimization decisions. This integration addresses the challenge of viewing compute and database costs together, offering a single view of combined costs for better resource allocation. The support includes daily updates of cost data from PlanetScale and allows for filtering by organization, service, cost category, and resource. The integration is available to all Vantage users without additional costs, although PlanetScale expenses will contribute to quota tier limits.

Analyzing AWS cost data is fundamental for effective cloud cost management, especially when planning, budgeting, or forecasting. AWS Cost Explorer offers detailed cost reports, but understanding its various categories, such as EC2-Other, and cost views—blended, unblended, and amortized—can be challenging. The EC2-Other category aggregates several unrelated subcategories, complicating cost forecasts, while the different cost views provide nuanced perspectives on expenses, with unblended reflecting real-time costs, amortized spreading upfront payments, and blended offering an averaged view. Discounts like AWS Enterprise Discount Plans, Savings Plans, and Reserved Instances are crucial for organizations with substantial cloud expenditure, providing significant savings. Tools like Vantage can enhance data analysis by reallocating costs to specific resources, offering a clearer picture of expenses and enabling accurate forecasting. Monitoring per-resource costs over extended periods aids in optimizing resource allocation, anomaly detection, and budget planning, beyond the default 14-day view in Cost Explorer, essential for anticipating future demand and optimizing cloud resources.

Amazon Aurora and PlanetScale are both MySQL-compatible database services that enhance performance and scalability while offering different pricing models. Amazon Aurora, compatible with MySQL and PostgreSQL, provides high performance and extensive customization options, including serverless configurations, but its pricing structure can be complex with numerous instance types and additional costs. PlanetScale, on the other hand, offers a straightforward and predictable pricing model, leveraging Vitess for scalability, with plans that include features like horizontal sharding and read replicas. In various scenarios, PlanetScale demonstrated cost savings over Aurora due to its inclusive pricing model, particularly in high-availability environments. While Aurora's versatility allows for tailored configurations, it can lead to unexpected costs if not carefully managed. PlanetScale's simplicity and predictability make it an appealing choice for users seeking consistent pricing without complex configurations.