Reverse bill all traffic along an IP address.
The Datafree Direct service is extremely low latency, secure (traffic is fully encrypted in transit), redundant, and highly scalable, with regional failover. There are two different versions of Direct: D-Direct and S-Direct.
D-Direct allows a customer to forward TCP (HTTP/S) traffic to an endpoint domain or static IP address. Forwarding traffic to a domain gives the customer the flexibility to manage their own domain resolution.
If the customer supports Proxy Protocol, D-Direct can pass the client/device IPs to the endpoint (origin server). In the case of unencrypted HTTP traffic, we add the device IP to the X Forwarded-For header.
By default, only ports 80 and 443 are open on the D-Direct service.
Going Live on D-Direct
To configure a D-Direct service, we need to know:
- The client domain – This is the domain that is being called from the client device.
- The endpoint domain or IP address – The endpoint can either be a customer-owned domain *that is different from the client domain above or a static IP address.
- Whether or not you will send a Proxy Protocol Header to us
- Whether or not you want us to send the Proxy Protocol Header to your endpoint – The reason for needing the Proxy Protocol Header is to be able to see the client/device IP address.
Prior to using the Datafree D-Direct service, a request from the Mobile Phone goes directly to the endpoint.
Once the information above is gathered, the D-Direct service can be provisioned. After provisioning, it is ready for traffic.
In order to go live with D-Direct, at which point the traffic will be datafree, the client domain must simply be pointed to the D-Direct service with a CNAME entry, as below.
Once your CNAME entry propagates, requests to app.customer.com now go to the D-Direct service before being forwarded to the Endpoint 1234.azurewebsites.net.
Your traffic is now datafree.
Please note, if your endpoint is defined as a static IP address instead of a domain, you will need to add an A Record instead of a CNAME.
S-Direct is similar to D-Direct in that it reverse-bills traffic to an endpoint. It differs, however, in that the endpoint must be a static IP address.
S-Direct supports TCP, UDP and GRE protocols. All ports/protocols are blocked on the S-Direct service except those specified in the setup. Any individual ports or port ranges can be opened on your dedicated S-Direct service.
Going Live on S-Direct
To configure an S-Direct service we need to know:
- The endpoint (static) IP address – All traffic that comes to your S-Direct service will be forwarded to your endpoint.
- Ports/Protocols you want open – All ports remain blocked on your S-Direct service except the ones you specify in the setup.
- Client domain (optional) – We like to know the client domain that is datafree so that we can monitor and troubleshoot when necessary. Technically, we do not need this for the S-Direct setup.
Prior to using the Datafree S-Direct service, a request from the Mobile Phone goes directly to the endpoint.
With the above information, the S-Direct service is provisioned and ready for traffic.
As with D-Direct, to go live with S-Direct, the client domain must be pointed to the S-Direct service with a CNAME entry, as below.
Once your CNAME entry propagates, requests to app.customer.com now go to the S-Direct service before being forwarded to the Endpoint IP address 18.104.22.168.
Your traffic is now datafree.
Alternatively, an A Record can be used instead of a CNAME. However, please note that using an A Record may result in infrequent, short (1-2 minute) outages when we push updates which require reprovisioning of the service.
*By customer-owned domain, we mean a domain for which the customer has name server access