Akamai is adding new features to its Ion network acceleration technology. The Ion product first debuted back in 2012 under the name Aqua Ion and is described by Akamai as providing situational performance for network optimization.
M.J. Johnson, director of product marketing at Akamai, explained to Enterprise Networking Planet that product names across all of Akamai’s solution families were simplified earlier this year.
“We identify our releases based on the timeframe when they were released,” Johnson said. “We’ve had two releases per year since launching Ion in 2012.”
Among the features set to debut is automated image compression with the JPEG2000 and WebP image formats.
Another new capability is something that Akamai refers to as Zone Apex Mapping, which can accelerate DNS. Johnson noted that Zone Apex Mapping is not a replacement for existing DNS services, whether provided by OpenDNS or another first or third-party DNS service.
“Zone Apex Mapping is a feature supported by Akamai that now allows domains at the so-called ‘apex of the zone’ (i.e. mycompany.com rather than www.mycompany.com) to be resolved and delivered by Akamai, rather than having to respond with an IP address to resolve outside of Akamai,” Johnson explained. “This helps avoid HTTP redirects and reduces DNS resolution time to save time resolving domains at the apex of the zone.”
Going a step further, Akamai is also testing out new features in its beta branch for Ion, which it refers to as the Innovation Channel. One of the new features that users can try out from the Innovation Channel is an algorithmic predictive pre-fetching capability.
Johnson explained that there are effectively two parts of the predictive pre-fetch capability. The first part is about studying and analyzing the traffic.
“For a short period after serving an html item (or other parent/referer) to the client, the edge ‘observes’ and studies which dependent objects are typically requested next,” Johnson said. “URLs of dependent objects are scored based on their likelihood to get requested following a request for the URL of the parent object.”
The second component of the algorithmic predictive pre-fetching capability is the prefetch decision itself. Johnson explained that when the edge network receives an html (or other parent/referrer) request from the browser, it will immediately prefetch the most likely dependent resources based on the study conducted in step 1.
“The edge does not wait to receive the actual html/parent/refer, but acts upon the URL and the objects scored to be highly likely to be requested by the client eventually,” Johnson said.
Sean Michael Kerner is a senior editor at Enterprise Networking Planet and InternetNews.com. Follow him on Twitter @TechJournalist.