Well-Architected Framework Series – Part 2 – Performance Efficiency – Design your cloud to achieve performance efficiency

Share on facebook
Share on twitter
Share on linkedin
Share on whatsapp

Well Architected Pillar: Performance Efficiency

The IT sector has been transitioning with cloud at high-speed, constantly demanding innovative techniques to keep up with the growing business requirements.

New designs, architectural constructs, software features and services are being developed and implemented divergently, to modernize businesses and enhance reliability, security and performance on cloud.

The AWS well-architected framework furnishes consistent improvement methods to build a potent architecture that can design and operate workloads on the cloud.

The framework is based on five pillars: Security, Performance Efficiency, Cost, Reliability and Operational Excellence.

In this article, the 2nd in the 5-part series of the pillars of the well-architected framework, we will be focusing entirely on the performance efficiency pillar.

The performance efficiency pillar utilizes computing solutions skillfully to meet the new-age tech surge while maintaining the efficacy of businesses.

Presiding cloud service providers like AWS, Azure, and GCP warrant best practices in design, delivery, and maintenance via their performance efficiency pillar. Organizations are able to better manage their production environments, scale workloads as per future demand, monitor resource availability, and provide additional infrastructure capacity when required.

Design Principles for Performance efficiency

The key principles that help maintain productive workloads on cloud are:

  • Embrace technological advancement: Formalize and entrust technology implementation to cloud vendors instead of burdening the IT teams to host and run new and complex technologies. Technologies like machine learning and NoSQL databases that need dedicated experts can be offered as cloud services by vendors, allowing teams to focus more on product development and innovation.
  • Become data-driven: Instrument business analytics and tools to create a valuable and secure data-driven culture that can provide timely insights across the organization at any given point in time.
  • Define patterns: Performance anti-patterns like applications unable to handle live workloads on production, scalability issues, and aligning with workload goals need to be identified and resolved to further convert these solutions into defined practices for better performances.
  • Testing: Perform load tests to ensure applications scale up when needed and do not underperform or collapse during peak traffic. Conduct comparative testing of instances, storage and configurations for different virtual and automated resources on cloud.
  • Use serverless architecture: Minimizes operational overheads and eliminates maintenance of physical servers enabling significant performance competence.
  • Monitor: Strengthen monitoring and control of new and current infrastructure and application services to ensure health and quality of workloads are relentless. Applying metrics and monitoring strategies to keep a check on scalability, flexibility and performance would enable effective optimization of resources.

Essential best practices for performance efficiency in the cloud

  1. AWS

AWS offers multiple resources and services to build SaaS and PaaS solutions that particularly focus on:

  • Select and configure specific cloud resources pertaining to compute, storage, database, and networking for higher performance.

Amazon EC2 provides a mix of optimal instances such as CPU, storage, memory, and networks that are correctly provisioned to match workloads, known as right-sizing. For example, AWS Lambda allows choosing the amount of memory needed for a function to deliver performance efficiency. AWS spot instances are used to generate loads and discover bottlenecks before moving workloads to production.

Key AWS database services include Amazon RDS, which help in simplifying operations and scaling databases, Amazon DynamoDB Accelerator (DAX) provides database performance auto scale support, Amazon Redshift scales SQL operations, Amazon Athena runs queries on data lakes with large amount of anomalous data and Amazon ElastiCache functions as an in-memory data cache to support applications that need high-speed responses.

AWS networking services are available in various forms depending on regions of deployment, data location and security and compliance requirements.

Amazon EBS optimized instances offer advanced configuration stack to minimize network conflicts while Amazon S3 transfer acceleration is for external users to benefit from uploading data to Amazon S3. AWS CloudFront, a global CDN, provides high-performance network connectivity to end-users, Amazon Elastic Network Adapters for network capacity for instances, Amazon Route 53 offers latency-based routing, Amazon Virtual Private Cloud (Amazon VPC) protects instances from internet traffic improving performance and AWS Direct Connect thatprovides dedicated connectivity to the AWS environment.

  • Once the approach is determined, an automated process to review and monitor AWS architecture performance is implemented.

Assess architectures on AWS by building technical as well as business performance measurement metrics and standards to benchmark and monitor data warehousing workloads.  Amazon CloudWatch, a monitoring service notifies violation if any and provides data analysis to evaluate performance, optimize resource utilization and gain insight on healthy and integrated operations.

A cloud solution test automation framework runs performance test scripts based on predictive analysis via the CI/CD process to evaluate performance efficiency vis-à-vis the expectations, track all test results back to the build and maintain a performance monitoring dashboard.

AWS Trusted Advisor helps monitor EC2 utilization, service limits and security as well as performance while optimizing AWS infrastructure and costs.

  1. Microsoft Azure

Azure PaaS offerings have the ability to render automated capacity scaling depending on workload peaks. Azure Autoscale services enable scaling via predetermined schedules whereas Azure Cache for REDIS is a sophisticated caching service that provides in-memory storage for faster retrieval of data. Azure Bob Storage creates data lakes for analytics and provides storage to build powerful cloud-native and mobile apps. Azure Monitor and Azure Marketplace solutions help monitor, test, and tighten the application performances. Similar to AWS, Azure does provide a plethora of services for specific purposes and provides Azure Advisor to help monitor and adhere to best practices for performance efficiency.

  1. GCP

In GCP, Compute Engine autoscaling allows the scaling of stateless apps on multiple identical VMs while the cluster feature in the Google Kubernetes Engine autoscaling caters to changing demands of workloads. Cloud Monitoring provides metrics and schedules for queue-based workloads and serverless compute options include Cloud Run, App Engine and Cloud Functions, each providing microservices autoscaling capabilities. Dataproc and Dataflow help scale data pipelines and data processing. Application performance management tools reduce latency and cost and Cloud TraceCloud Debugger and Cloud Profiler help gain analysis and remediation into how code and services function. As more and more services get provisioned GCP is evolving in making performance efficiency as one of the key pillars and area of advantage over other cloud providers.

How to enhance Performance efficiency on cloud?

The complex cloud with multiple components at varied locations and limited monitoring strategies, make it difficult to ensure if applications perform and communicate optimally.  To reduce or eliminate latency and unpredictable performances on shared devices and identify problematic areas, it is important to attain high performing cloud environments. To enhance performance:

  • Use the right instances. Cloud providers & users must right size and provision for appropriate instances that match business needs while enhancing workloads.
  • Implement load balancing and autoscaling services in the dynamic cloud to monitor and correct issues, if any, with respect to workloads, database and resource utilization. It helps gather metrics to improve the overall health of the cloud.
  • Ensure data resides in the right place to avoid bottlenecks and minimize the need to keep moving it between environments.
  • Use microservices optimally. The small independent components of microservices work via APIs that may tax or limit performances distinctly.
  • Utilize multiple caches through content delivery networks for speedy access to data storage.
  • Eliminate deployment and operations of perpetual VMs or container instances and adapt to event-driven architecture that can be run on the serverless cloud.
  • Implement automation and orchestration tools to minimize manual efforts, speed up tasks and enable error-free functioning and support of cloud operations. This is where tools like CloudEnsure come in handy to improve performance efficiency.


Sustaining and optimizing performance while keeping up with the technological evolution depends mainly on identifying which workloads & what configurations affect performances to what extent and whether they are a priority. Today the customer wins are being defined with small margins and those margins are a direct result of how efficiently your complete infrastructure is setup. Based on the review analysis and metrics, organizations must determine appropriate resources and solutions to be more data-driven, build the right architecture and drive eminent monitoring and evaluation practices to augment business performances. The sooner one understands how it has a direct impact the better gains one can make and sustain the business.

Share on twitter
Share on linkedin
Share on facebook
Share on whatsapp

Leave a Comment

Your email address will not be published. Required fields are marked *