New fault action in AWS FIS to inject I/O latency on Amazon EBS volumes

Today, Amazon EBS announced a new latency injection action in AWS Fault Injection Service (FIS), a fully managed service for running fault injection experiments. You can now use this action to inject I/O latency on your volumes as part of a controlled testing experiment to understand how your mission-critical applications respond to storage faults. With the new fault action, you can test your architecture against elevated storage latency, allowing you to observe application behavior and fine-tune your monitoring and recovery processes to ensure high availability.

EBS volumes are designed to meet the needs of highly available, latency-sensitive applications such as Oracle, SAP HANA, and Microsoft SQL Server. The latency injection action simulates degraded I/O performance on your volume to replicate the real-world signals, such as Amazon CloudWatch alarms and operating system timeouts, that occur during storage performance issues. Using this action, you can build confidence that your application can withstand and quickly recover from disruptions that cause high I/O latency on your EBS volume. To get started, you can directly use the pre-defined latency injection experiment templates available in the EBS and FIS consoles. Alternatively, you can customize these experiment templates or create your own experiment templates to meet your application-specific testing needs. You can integrate these latency injection experiments into your existing chaos engineering tests, continuous integration, and release testing, as well as combine multiple FIS actions in one experiment.

This new action is available in all AWS Regions where AWS FIS is available. To learn more, visit the EBS FIS actions user guide.