One of the key requirements for any Openstack cloud deployment is High availability(HA). High Availability refers to the availability of resources in the wake of component failures in the system. Hardware or software failures, bugs, traffic spikes combined with a lack of redundancy can bring even the best-resourced Openstack cloud crashing down. Even if applications are architected for resiliency, these efforts are futile if critical components suffer an outage. Before deploying your public and private Openstack clouds in production, how do you validate that they are highly available? How can you assure that your Openstack cloud can sustain server-level outages?
In this session, join Behzad Dastur, Ajay Kalambur, and Pradeep Chandrasekar, as they cover High Availability testing strategies at individual component level and system level. Geared for anyone who wants to get an idea of various test strategies for Openstack HA , attendees will learn:
- How to test OpenStack for resiliency with concurrent load operations during an availability disruption event
- Different types of HA triggers across process, system, and network levels
- Limitations in existing OpenStack testing tools such as Rally and Tempest from a HA test perspective
- Overview of the HA tool we developed to automate HA testing
- Testing of database nodes, RabbitMQ, compute and network nodes for HA
- Use cases and examples of tailored configurations for HA
Come learn about HA testing and accompanying tools so you can replicate this in your Openstack cloud.