One important aspect of each Infrastructure as a Service offering is the performance of the infrastructure. At first glance, the offers on the market seem to be reasonably comparable – everyone offers cores, RAM and disk space. But even though the parameters appear to be approximately the same on paper, there’s no automatic guarantee that you’ll really get the same performance. They can be worlds apart. ProfitBricks’ infrastructure is on average twice as fast as that of the main competitors. And it’s only half as expensive. Sounds unbelievable? Test it with your own workload. Completely free of charge.
Performance Tests With Standard Benchmarks
So, although they cost less, the ProfitBricks servers are faster than those of Amazon and Rackspace. Cloud Spectator has recently proven in benchmark tests that ProfitBricks is twice as fast. Take a look here at the results of the tests against the benchmarks DBench, UNIXBench and iPerf.
Why Do Such Differences Exist?
The basic technology of IaaS providers is entirely different. While some providers, for example, use systems that were developed 5 or more years ago, ProfitBricks uses custom-developed network tools that comply with state of the art technology.
Many providers also use hardware systems that are already held in the company’s portfolio and have been written off (such as discarded “dedicated servers”). They thus avoid the high expenditure of replacing all the existing software with more modern systems – and sell you their old systems at a high cost. The ProfitBricks data centers, on the other hand, are equipped with up-to-date hardware specifically aligned to IaaS applications. The outstanding price/performance ratio of ProfitBricks’ infrastructure proves that this can still be favorably priced – twice as much performance, half the price.
In addition to the technology and the hardware, the performance of an IaaS provider also depends on the application you use. Depending on the operating system, programming language, framework or database, performance bottlenecks can occur in certain performance subareas of your application, so that you can expect losses in performance over a longer period of time from some providers.
Testing IaaS Objectively Under Real Conditions
To be really certain of what performance you can expect from a particular IaaS provider, it is recommended to install your application on a test system and to measure the speed objectively. This is the only way to determine performance properly and to compare providers‘ prices realistically.
Checking Continuity Of IaaS Performance
Some providers’ performance can fluctuate considerably over time. If the instances are used by several customers, for example, your performance might be restricted from such a provider if other customers generate a high level of load. ProfitBricks provides you with the IaaS infrastructure performance exclusively and reliably in dedicated instances.
Our recommendation: You should always test the performance of your system over a longer period of time, so that you can effectively assess the continuity of the service performance.