Benchmarking the Infrastructure – Evaluating Resource Capacity and Optimization

“Save for a rainy day.”

– Aesop

One of the major challenges when operating a growing infrastructure is keeping different services up and running within the expected defined Service Level Agreements (SLAs). During the course of production, unexpected issues might arise, even though a proper monitoring solution has been put in place to act proactively when certain situations are detected, such as scaling cluster nodes up and down to accommodate more tenant workloads. In a complex environment such as OpenStack, hitting system performance limits is considered a common issue due to its distributed and loose architecture. The more services join the ecosystem, the more performance issues will be raised. You might have a monitoring dashboard where all services are showing green but cloud tenant users are facing issues with spawning a Virtual Machine (VM) after several failed requests...