-2

I was creating client to insert, update and read data to geomesa accumulo storage. I have gone through many of the documents like using thrift proxy and using WFS-T standard implemented by geoserver. Using thrift proxy creates lots of abstraction and i am not able to understand the communication between accumulo and client clearly, whereas on the other hand WFS-T does provides http service for transactions but uses xml payload data in request body.

Both this techniques are implemented using java framework which uses accumulo jar files to do rest.

I want to understand how can i achieve same thing in other language using the base protocols used by one of the technique?

What can be done to achieve same thing without being dependent on java?

Suresh Prajapati
  • 3,991
  • 5
  • 26
  • 38
  • "How exactly accumulo internal works?" Try to keep your questions more specific. These are too vague to result in a good answer. I've tried to provide a high-level answer below, but you'd do better to improve your questions on StackOverflow in the future. Also, you mention geomesa, but this appears to be a question about Accumulo and it's not clear what geomesa has to do with it. – Christopher Mar 09 '17 at 04:01

1 Answers1

2

Apache Thrift is a framework for producing custom client-server protocols on common design principles. It comes with code generation tools and multiple language-specific libraries to help you create your protocol.

The native API for Apache Accumulo is Java. However, Accumulo also provides the "thrift proxy" that you mentioned. This is a Java application which listens for client connections to its Thrift service, and proxies the commands to Accumulo using Accumulo's Java API. This lets you write your actual client code in any Thrift-supported language.

Both Accumulo and Thrift are open source projects, with community mailing lists. You should visit their websites and their respective mailing lists for more details.

Christopher
  • 2,427
  • 19
  • 24