As virtualization options proliferate in the era of massive cloud computing and agile infrastructure, you face important choices that are likely to have long-term effects on your computing and data handling performance.
The question is this: How do you determine “real” performance differences?
We worked with independent cloud analyst agency Cloud Spectator to assess several competing virtualization solutions, including a commercial hypervisor, CentOS KVM , and Virtuozzo Containers™.
Within each platform, Cloud Spectator deployed equivalent load generators to saturate the physical host and emulate resource contention on a production environment. Cloud Spectator then used one VM/container to run benchmark tests to observe performance and performance variation of the loaded environment. The testing was conducted over a 24-hour period, excluding setup time for the physical server, hypervisor installation, and load generation.
The report yields some interesting insight in how containers compare to hypervisor-based solutions when deploying virtual private servers. You might be surprised at the differences in some areas, including processor performance.
For complete details of the test methodology, results and conclusion, get the report here and see what makes the most sense for your VPS deployments.