As more and more companies begin offering cloud-based services and, in turn, as more and more companies begin to migrate to the cloud, there’s an increasing demand for tools to monitor and assess cloud performance. Although we hear a lot of about security in the cloud, a study released late last year by the market research firm IDC listed “performance” as one of IT’s major concerns, ahead of cost and vendor lock-in.
Last week, we looked at CloudFail.net, a blog that tracks the RSS feeds of major cloud providers in order to monitor service updates and outages. But a number of other services exist that can help customers assess the dependability of cloud providers.
Developed by Compuware Gomez, CloudSleuth is a cloud performance visualization tool initially created as an internal resource to help us gauge the reliability and consistency of the most popular public IaaS and PaaS providers. CloudSleuth uses the Gomex Performance Network to measure the performance of an identical sample application running on popular cloud service providers, assessing two basic user experience metrics – response time and availability. The tests are currently run from locations in all 50 states and from 75 international locations, and there are plans to add the ability to benchmark a user’s own application.
While much of CloudHarmony still in beta, it looks to become an important resource for evaluating performance. Currently, you can use its Cloud SpeedTest to test upload and download speeds, page loads and latency on several major services. The CloudHarmony blog also contains a number of analyses of various services, including encoding, CPU performance, and memory I/O.
Cloudstone is a multi-platform, multi-language performance measurement tool for Web 2.0 and Cloud Computing. This UC Berkeley project is described as “a toolkit consisting of an open-source Web 2.0 social application (Olio), a set of automation tools for generating load and measuring its per- formance in different deployment environments, and a rec- ommended set of constraints for computing a metric we believe makes more sense, dollars per user per month.” While the results from the project aren’t published, the developers have created an application that gives users the ability to do the research themselves.
Developed by Duke University and Microsoft Research, Cloud CMP “pits cloud against cloud,” assessing computation, storage, and network services offered by different cloud providers, then estimate performance and cost of an application if it’s deployed on a particular cloud provider.
Do you know of other resources out there to help gauge cloud performance? And what other assessments should tools like these be making?