Faced with the problem of measuring various workloads we used OpenStack Infra.
Few of these required a long running OpenStack deployment with constant workload resembling real world scenarios.
Our initial focus was to attempt to emulate these workloads using a number of tools such as Tempest, Heat Stacks, Rally, User Stories, etc... It becomes difficult for any team, both small and large, to tackle this task and provide quality feedback without making this another project on its own. We thought why even simulate a workload when we can have a real thing?
With Big Tent, OpenStack Infrastructure requirements for computing resources grows with every project or developer. That means they are in constant need of resources. In this talk we would like to present how we used OpenStack Infrastructure workload to measure API uptime and VM performance of our OpenStack deployments and helped OpenStack community at the same time.