Apigee Hadoop integration

[#1]

I am trying to understand whether Apigee provides a container to develop restful APIs to a Hive Data store. I have orders data in Hive and my RESTful API needs to provide a getOrderStatus<OrderId> support. All the data needs to come from Hive in realtime. Can you point me to the technologies and software components that I need to use within Apigee.

[sarthak · Accepted solution]

Hi @Staran you can connect to any backend which is exposed via HTTP using Apigee EDGE. You can do it using the proxy wizard.

If the backend is non HTTP you can connect to it using node.js . I found the following node packages for connecting to Hive.

https://www.npmjs.com/package/node-hive

https://www.npmjs.com/package/thrift-hive

And there were a few others as well :: https://www.npmjs.com/search?q=hive

If you don't know how to use node.js in Apigee you can start from here.

Sarthak

[#3]

Are you saying we will need to write a thrift client (using node.js) and also a thrift server to connect to Hive. I am trying to understand where we will be writing the implementation of the GET www.mycompany.com/orderStatus<orderId>; REST call. There is business logic and I want to know whether that is written in the apigee ecosystem or outside.

Secondly, does Apigee manage connection pooling on backend or does each call create a new connection to get a Hive handle.