February 2015 Web Host Performance Update

Cloud Spectator released the CloudSpecsTM Web Host Monitor in order to help consumers make informed decisions when selecting web hosting providers and promote transparency in the shared hosting space. By tracking the uptime and response time of twenty-four Linux, Windows and Managed WordPress offerings, we were able to present performance rankings based on data driven results.

Below is a summary of the uptime and response time results of a variety of web hosting providers from the CloudSpecsTM Web Host Monitor.

Summary

Linux providers altogether showed average response time among all three offering types. Among Linux hosting providers, A Small Orange exhibited the shortest response time (535 ms), and GoDaddy the second at 625 ms. 1&1 and BlueHost exhibited the highest Linux uptime result in December (100%), followed by A Small Orange, HostGator and iPage with 99.99% uptime results respectively.

Compared to the results from Linux and Managed WordPress hosting options, the response times of Windows offerings showed significantly larger variations, with the response time of the slowest provider being 34 times as long as the fastest one. This trend has lasted since August 2014. Rackspace had the best response time (401 ms), followed by HostGator and GoDaddy at 548 ms and 558 ms respectively. HostGator also showed 99.97% uptime in February, and 1and 1 and Rackspace displayed 99.89% uptime results.

Generally speaking, Managed WordPress had the shortest average response time compared to Linux and Windows hosting options. Its average response time was around 2 times shorter than the Linux offering and nearly 10 times shorter than the Windows offerings (due to the large variation among Windows providers). FlyWheel, Pagely, WPEngine, BlueHost, and GoDaddy Managed WordPress offerings displayed the shortest response time among all hosting options (224 ms, 283 ms, 313 ms, 322 ms and 352 ms respectively). FlyWheel showed 100% uptime in February, followed by BlueHost with a 99.99% uptime.

It should be noted that although the data collected on uptime and response time is a good indication of web hosting performance, users should be cautious when generalizing those results given that the tests were performed in all default settings with limited location choices. If any specific hosting requirements are considered, we suggest customers go through relevant consultations and perform customized tests.

Uptime Leaders

Higher percentage is better, as it indicates the amount (in percentage) of overall time the server has been up for the month.

Response Time Leaders

Lower values are better for response time as they indicate less latency.

Methodology

CloudSpecsTM gathered accurate metrics for response time and uptime with Pingdom’s monitoring service. Cloud Spectator adopted Pingdom’s methodology for measurements to further maintain objectivity.

Cloud Spectator tracked three types of hosting offerings: Linux, Windows and Managed WordPress. For each offering, a mock website was created to simulate end user experience as accurately as possible. Linux websites used WordPress, while Windows websites used DotNetNuke (DNN). All web content (images, text, etc.) were hosted on the local server, thus performance was not dependent on files or objects stored outside of the web server. Not all of the providers carried all three types of offerings; therefore, the providers listed in each section vary.

The web hosting providers measured in the three offerings were:

Cloud Spectator set up anonymous accounts on each hosting provider using all default settings.

Pingdom gathered uptime data by pinging the server at set intervals. The web server was recorded as being up if a response was received. If no response was received for two sequential pings, the web server would be marked as having a down time during that interval.

Pingdom gathered response data when the server was up. Response time was tracked using Pingdom’s Node Group, a reserved feature for its Enterprise Users. The node pinged the web server and recorded the time duration between when the message was sent and when it was received back as an indication of network speed. The time durations were tracked globally from several locations in a node group on a rotational basis, which included the following probes: