OnApp CDN in-a-box enables companies to become a CDN specialist, not a generalist on the same level of Akamai, Limelight or EdgeCast. The CDN specialist occupies a certain place in the CDN ecosystem. The OnApp CDN software is not a fit for every company, but only for a select few. OnApp CDN does work well for hosting companies and Telco’s, that offer CDN services as an option, alongside a bucket of other services, such as cloud compute and cloud storage. OnApp is not a fit for the pure-play CDN that offers all the CDN bells and whistles.
Pure-play CDN
CDNify is a pure-play CDN whose entire livelihood depends on selling CDN services. The best scenario for a pure-play CDN is to build its infrastructure from the ground up, just like Fastly, Yottaa and Instart Logic. OnApp is not to blame for CDNify’s circumstances. Based on the available information, it seems that the expectations of CDNify were not in line with how the CDN ecosystem works. OnApp further exacerbated CDNify’s situation with its comments on how the CDN ecosystem works in their view. OnApp enables hosting companies to become like Internap, not compete with Akamai.
Infrastructure Consistency
From a technology standpoint, as every pure-play CDN understands, it’s advantageous to keep all hardware, software, and ISPs used across all POPs consistent. Keeping the same ISP across all POPs, whether it’s Level 3, or BT, or whoever, enables the CDN to optimize routing between POPs, especially when it comes to middle mile optimization. Middle mile optimizations requires control of all end points (POPs), especially if a CDN is using anycast. Problems arise if each POP is using a different ISP. OnApp works against the CDN consistency approach, since its entire ecosystem of hosting partners all use different brands of servers, routers and ISPs.
Conclusion
CDNify will have to build its CDN using Varnish or Nginx, then deploy the same brand of servers and routers across all POPs, using the same ISPs for bandwidth. This of course requires capital. If the company doesn’t have the capital to build out a 10 POP CDN, than it shouldn’t get into the CDN business, unless they have money to burn. In regards to comments made by CDNify about competition, they are absolutely correct. The CDN industry is fiercely competitive. That is why CDNify needs to become a specialist in one or two areas within the CDN ecosystem, and be the best it can in that niche, whether its security, user engagement, time-to-first paint, or whatever. Part 2, Akamai and Limelight vs OnApp