A curator service discovery server implementation
#Requirements
- git
- jdk 1.7
- mvn 3
- zookeeper 3.4.6
#Install
Clone the project and install it!
$ git clone https://github.com/kpacha/suricate.git
$ cd suricate
$ mvn install
And the suricate-0.0.2-SNAPSHOT.jar should be on your local maven repo
#Environment
Start your ZooKeeper server(s) before running suricate. More details on http://zookeeper.apache.org/doc/trunk/zookeeperStarted.html
#Run with maven
If your ZooKeeper is running in your local machine with the default config, just type one more maven command and done!
$ mvn clean compile exec:java
Your service should be waiting for you at port 8080. Nice, uh?
#Build and Run the fat-jar
So, let's build it for real, deploy it to an actual server and run it. Remember, in production environments, you should run your ZooKeeper services in replicated mode...
$ mvn clean compile assembly:single
# ...and you are ready for deploy the fat-jar! (it's placed at `target/`)
# ... your deployment process here ...
# start the suricate service
$ java -jar suricate-0.0.2-SNAPSHOT-jar-with-dependencies.jar [-p <jetty_port>] [-b <zk_base_path>] [-c <zk_connection_string>] [-t <instance_cleanup_time>]
And you already have a suricate server running!
The options are:
-p <arg> jetty port (default '8080')
-c <arg> zookeeper connection string (default 'localhost:2181')
-b <arg> zookeeper base path (default '/suricate/service-directory')
-t <arg> instance cleanup time in millis (default '15000')
#Usage
Register your service nodes (hosts and clusters) by sending a PUT request to the putService method
# Register the host 'ca2fff8e-d756-480c-b59e-8297ff88624b' in the cluster 'test'
$ curl -i -H "Accept: application/json" -H "Content-Type: application/json" -X PUT -d '{"name": "test", "id": "ca2fff8e-d756-480c-b59e-8297ff88624b", "address": "10.20.30.40", "port": 1234, "payload": "supu", "registrationTimeUTC": 1325129459728, "serviceType": "STATIC"}' http://localhost:8080/v1/service/test/ca2fff8e-d756-480c-b59e-8297ff88624b
In one or two instance cleanup cycles the garbage collector will remove this STATIC service node. Set the serviceType as PERMANENT for persistent nodes.
Get a list of registered services with the getAllNames method
$ curl -i http://localhost:8080/v1/service
Get a list of registered nodes in the 'test' service (hosts in a cluster) with the getAll method
$ curl -i http://localhost:8080/v1/service/test
Get a registered node in the 'test' service by its id with the get method
$ curl -i http://localhost:8080/v1/service/test/ca2fff8e-d756-480c-b59e-8297ff88624b
Remove a service node with the removeService method
$ curl -i -X DELETE http://localhost:8080/v1/service/test/ca2fff8e-d756-480c-b59e-8297ff88624b
Check the curator-x-discovery-server for more details about the Discovery Service and its REST interface.
The REST entities are described in the curator-x-discovery-server repo