Does APIGEE makes HTTP Call to the Target Endpoint?

Once I create a Proxy lets say I am saying

<APIGEE_HOST>/sendRequestToAPIGEE

this Request originally maps to

https://www.google.com

Now once I do a GET on <APIGEE_HOST>/sendRequestToAPIGEE , the request hits the APIGEE servers and then passes though the proxies I have written and in the end APIGEE creates a GET request to http://www.google.com and returns the response to me?

Is this what happens? Or APIGEE creates a Tunnel internally and use this tunnel everytome?

0 5 443
5 REPLIES 5

@Unicron.. No APIGEE does not create any tunnel , its a pure HTTP/s call from APIGEE edge message processors to southbound target...... In your case google.com

Hope this helps.

@Pratik Gupta Thanks a lot. It would a costly operation for APIGEE edge servers then. Costly in terms of Bandwidth,Response Time. If you guys run on JVM (which is what I assume) then do you guys run on non-blocking IO ? Or you spawn a Thread per request ?

We don't describe the insides very much. The implementation is not part of the documented interface of the service.

Anyway what's the source of your interest?

@Dino-at-Google my source of Interest is my curiosity. Currently its a black box to me. I use it in my life but I don't know how it works underneath -is a shame to myself! If I know the core I would be able to utilize the software to the fullest. The following very basic questions would help

  • What is the Threading Model used underneath or its Event Loop?
  • Does APIGEE depends on Netty or JAVA NIO ?
  • Why JAVA? The problem solved by APIGEE has a very valid use case for NODEJS -for asynchronous reasons .

It's supposed to be a black box!

We don't document the internals, or describe them. They are subject to change. We also don't answer design questions like "why Java" beyond obvious generalities like "that's the language we were comfortable with, when we built this thing!"

If you have questions on the documented interfaces, I'll be glad to direct you.