Menu
CDN Pro is an edge cloud platform with a server group-based price model that offers price transparency. The price model is ideally suited for the modern CDN and removes uncertainties from the billing process. We organized our global Points of Presence (PoPs) into server groups based on cost.
We organized our global Points of Presence (PoPs) into 4 server groups based on cost. From lowest to highest, the groups are:
Depending on your needs, you can choose the server groups to use for your visitors according to country and ISP. This lets you maximize performance and minimize costs.
We also introduced a charge by CPU usage for the first time in the industry. This unified price model can accurately reflect the cost consumed to serve customers with different types of traffic. It also allows us to offer sophisticated edge computing features without sophisticated price matrix.
You can learn more technical details from this blog.
Assume your main objective is to ensure the performance of your website in Japan and Korea. In this scenario, you may want to configure the edge hostname to use all server groups outside China to serve these two countries while using the “standard” group to serve the rest of the world. The screenshot below shows the completed configuration on the portal.
At the end of the billing cycle, the invoice shows the amount of traffic served by each server group and the CPU usage with the associated charges with a minimum of $50.
CDN Pro’s global network covers the most significant areas on the planet and continues to expand the reach of its global network. Its points of presence (PoPs) are organized into “server groups” with a different price for traffic served from each group. You can choose which server groups to use for each country and ISP, allowing you to fully customize performance and cost based on your budget.
We organized our global Points of Presence (PoPs) into server groups based on cost. Depending on what you need, you can choose the server groups to use for your visitors according to country and ISP. This lets you maximize performance and minimize costs.
For consistency, server hardware is largely the same across server groups. The main difference between different server groups is the bandwidth cost charged by our vendors (ISPs). A PoP in the “ultra” group is more expensive than one in the “standard” group, but does not necessarily deliver better performance than a “standard” PoP. Performance experienced by each end user is determined by the connectivity between the user and the PoP. Adding the “ultra” group to serve a region just provides our GSLB algorithm more PoPs to choose from, such that the end users in that region are more likely to get a PoP with better connectivity.
No. The CDN Pro GSLB assigns PoPs primarily based on performance. Adding “ultra” to serve a region enables the GSLB to choose from a larger pool of PoPs for that region. If the algorithm determines that a request will be better served by a “standard” PoP, the traffic is still served from that PoP and incurs a “standard” charge.
In general, the “Standard” group delivers excellent coverage for North America , Europe, and some regions in Asia. Adding the “Premium” group can enhance the performance in those areas as well as the Middle East. Adding “Deluxe” delivers good performance in APAC countries. Adding “Ultra” gives you our best possible performance in all areas including South America, Oceania and Africa. We invite you to try different configurations to determine the right balance between performance and cost.
CDN Pro does not charge for TLS certificates and HTTPS accesses. You can BYO certificate for your property domain configuration. CDN Pro also supports self-signed certificates and auto-renew through Let’s Encrypt.
CDN Pro collects CPU usage based on NGINX’s “event-driven, non-blocking” architecture and calculates the total CPU time consumed to handle a request. For more technical details about the process of collecting CPU time, refer to this blog.
To determine the amount of CPU resources required to deliver a fixed amount of data, divide the total CPU time by the traffic volume. Based on our measurements of a set of domains containing different content, a domain with highly cacheable large file content can take as little as 1.56 seconds to transfer 1GB of data, while a domain with dynamic API service content can take 453.66 seconds. The results show significant differences in CPU time by domains serving different types of content.
Based on the per-CPU hour price of $1.95, the cost of CPU consumption for the above 1GB data transfer can be calculated as follows:
For highly cacheable large file content: (1.56s/3600) * 1.95 * 100 = 0.0845 cents.
For dynamic API service content: (453.66s/3600) * 1.95 * 100 = 24.57 cents.
