Best practices for designing and architecting with Amazon Keyspaces (for Apache Cassandra)

Use this section to quickly find recommendations for maximizing performance and minimizing throughput costs when working with Amazon Keyspaces.

Contents

• Key differences and design principles of NoSQL design

  • Differences between relational data design and NoSQL

  • Two key concepts for NoSQL design

  • Approaching NoSQL design

• Optimize client driver connections for the serverless environment

  • How connections work in Amazon Keyspaces

  • How to configure connections in Amazon Keyspaces

  • How to configure the retry policy for connections in Amazon Keyspaces

  • How to configure connections over VPC endpoints in Amazon Keyspaces

  • How to monitor connections in Amazon Keyspaces

  • How to handle connection errors in Amazon Keyspaces

• Data modeling best practices: recommendations for designing data models

  • How to use partition keys effectively in Amazon Keyspaces

    • Use write sharding to evenly distribute workloads across partitions

      • Sharding using compound partition keys and random values

      • Sharding using compound partition keys and calculated values

• Optimizing costs of Amazon Keyspaces tables

  • Evaluate your costs at the table level

    • How to view the costs of a single Amazon Keyspaces table

    • Cost Explorer's default view

    • How to use and apply table tags in Cost Explorer

  • Evaluate your table's capacity mode

    • What table capacity modes are available

    • When to select on-demand capacity mode

    • When to select provisioned capacity mode

    • Additional factors to consider when choosing a table capacity mode

  • Evaluate your table's Application Auto Scaling settings

    • Understanding your Application Auto Scaling settings

    • How to identify tables with low target utilization (<=50%)

    • How to address workloads with seasonal variance

    • How to address spiky workloads with unknown patterns

    • How to address workloads with linked applications

  • Identify your unused resources to optimize costs in Amazon Keyspaces

    • How to identify unused resources

    • Identifying unused table resources

    • Cleaning up unused table resources

    • Cleaning up unused point-in-time recovery (PITR) backups

  • Evaluate your table usage patterns to optimize performance and cost

    • Perform fewer strongly-consistent read operations

    • Enable Time to Live (TTL)

  • Evaluate your provisioned capacity for right-sized provisioning

    • How to retrieve consumption metrics from your Amazon Keyspaces tables

    • How to identify under-provisioned Amazon Keyspaces tables

    • How to identify over-provisioned Amazon Keyspaces tables