Data and storage design considerations for sustainable workloads on Azure

Data storage in Azure is a crucial component of most provisioned workloads. Learn how to design for a more sustainable data storage architecture and optimize existing deployments.

Important

This article is part of the Azure Well-Architected sustainable workload series. If you aren't familiar with this series, we recommend you start with what is a sustainable workload?

Storage efficiency

Build solutions with efficient storage to increase performance, lower the required bandwidth, and minimize unnecessary storage design climate impact.

Enable storage compression

Storing much uncompressed data can result in unnecessary bandwidth waste and increase the storage capacity requirements.

Green Software Foundation alignment: Hardware efficiency

Recommendation:

  • A solution to reduce the storage requirements, including both capacity and required bandwidth to write or retrieve data. For example, compressing files in Azure Front Door and compressing files in Azure CDN.
  • Compression is a well-known design technique to improve network performance.
  • Consider the tradeoff of compression: Does the benefit of compression outweigh the increased carbon cost in the resources (CPU, RAM) needed to perform the compression/decompression?

Optimize database query performance

Querying extensive databases or retrieving much information simultaneously can have a performance penalty. Ideally, apps should optimize for query performance.

Green Software Foundation alignment: Energy efficiency

Recommendation:

Use the best suited storage access tier

The carbon impact of data retrieved from hot storage can be higher than data from cold- or archive storage. Designing solutions with the correct data access pattern can enhance the application's carbon efficiency.

Green Software Foundation alignment: Energy efficiency

Recommendation:

Only store what is relevant

Backup is a crucial part of reliability. However, storing backups indefinitely can quickly allocate much unnecessary disk space. Consider how you plan backup storage retention.

Green Software Foundation alignment: Hardware efficiency

Recommendation:

  • Implement policies to streamline the process of storing and keeping relevant information. Microsoft Purview can help label data and add time-based purging to delete it after a retention period automatically. Additionally, this lets you stay in control of your data and reduces the amount of data to process and transfer.
  • Workloads integrated with Azure Monitor can rely on Data Collection Rules (DCR) to specify what data should be collected, how to transform that data, and where to send the data.

Determine the most suitable access tier for blob data

Consider whether to store data in an online tier or an offline tier. Online tiers are optimized for storing data that is accessed or modified frequently. Offline tiers are optimized for storing data that is rarely accessed.

Green Software Foundation alignment: Energy efficiency

Recommendation:

Reduce the number of recovery points for VM backups

Recovery points aren't automatically cleaned up. Therefore, consider where soft delete is enabled for Azure Backup. The expired recovery points aren't cleaned up automatically.

Green Software Foundation alignment: Hardware efficiency

Recommendation:

Revise backup and retention policies

Consider reviewing backup policies and retention periods for backups to avoid storing unnecessary data.

Green Software Foundation alignment: Hardware efficiency

Recommendation:

  • Review and revise backup and retention policies to minimize storage overhead.
  • Actively review and delete backups that are no longer needed.

Optimize the collection of logs

Continuously collecting logs across workloads can quickly aggregate and store lots of unused data.

Green Software Foundation alignment: Energy efficiency

Recommendation:

Next step

Review the design considerations for security.