Monday, September 9, 2013

Exploring Apache Camel Core - Seda Component

The seda component in Apache Camel is very similar to the direct component that I’ve presented in previous blog, but in a asynchronous manner. To do this, it uses a java.util.concurrent.BlockingQueue as default implementation to queue up messages and disconnect from your main Route thread and then processing the messages in a separated thread. Because of this BlockingQueue, you need to be aware of the usage and config option.

One option needs to be aware of asynchronous processing is the it default to queue size is unbound, meaning it will grow as much as your memory allowed. To cap this, set size=1000. Let’s see an example.

package camelcoredemo;

import org.slf4j.*;
import org.apache.camel.*;
import org.apache.camel.builder.*;
import org.apache.camel.main.Main;
import java.io.*;

public class SedaDemoCamel extends Main {
    static Logger LOG = LoggerFactory.getLogger(SedaDemoCamel.class);
    public static void main(String[] args) throws Exception {
        SedaDemoCamel main = new SedaDemoCamel();
        main.enableHangupSupport();
        main.addRouteBuilder(createRouteBuilder1());
        main.addRouteBuilder(createRouteBuilder2());
        main.addRouteBuilder(createRouteBuilder3());
        main.run(args);
    }
    // The file poller route
    static RouteBuilder createRouteBuilder1() {
        return new RouteBuilder() {
            public void configure() {
                from("file://target/input?preMove=staging&move=.processed")
                .process(new Processor() {
                    public void process(Exchange msg) {
                        CamelContext camelContext = msg.getContext();
                        ProducerTemplate producer = camelContext.createProducerTemplate();
                        String text = msg.getIn().getBody(String.class);
                        String fileName = (String)msg.getIn().getHeader("CamelFileName");
                        boolean specialFile = fileName.endsWith("_SPECIAL.dat");
                        if (specialFile)
                            producer.sendBody("seda:specialRoute", text);
                        else
                            producer.sendBody("seda:normalRoute", text);
                    }
                });
            }
        };
    }
    // The special file processing route
    static RouteBuilder createRouteBuilder2() {
        return new RouteBuilder() {
            public void configure() {
                from("seda:specialRoute")
                .process(new Processor() {
                    public void process(Exchange msg) {
                        LOG.info("Processing special file: " + msg);
                    }
                });
            }
        };
    }
    // The normal file processing route
    static RouteBuilder createRouteBuilder3() {
        return new RouteBuilder() {
            public void configure() {
                from("seda:normalRoute")
                .process(new Processor() {
                    public void process(Exchange msg) {
                        LOG.info("Processing normal file: " + msg);
                    }
                });
            }
        };
    }
}

You will notice that this demo code is very similar to the direct component demo, with few differences. First, we use seda endpoints. Second, in file poller, we read in the entire file content text. We do this because we are now passing to an asynchronous Route that will runs on separate threads. The poller is configured to move the processed file into different folder right after the first Route has ended. So we must ensure the processing Route is not depended on the path of the File, hence we will load entire text in instead.

Another interesting seda option is you may set the number of concurrent threads to receive the messages to process them! Let’s say if your normal files are heavy in traffic, then you can configure to use more threads on that part (default is just one thread.)

from("seda:normalRoute?concurrentConsumers=10")
.process(new Processor() {
    public void process(Exchange msg) {
        LOG.info("Processing normal file: " + msg);
    }
});

To verify that your are running concurrently, you can easily configure your logger to display thread name. For example with log4j, you can use this pattern:

log4j.rootLogger=INFO, stdout
log4j.appender.stdout=org.apache.log4j.ConsoleAppender
log4j.appender.stdout.layout=org.apache.log4j.PatternLayout
log4j.appender.stdout.layout.ConversionPattern=%d %p %t [%c] - %m%n

There are more options availabe from Seda component that you may explore. Try it out with a Route and see it for yourself.

No comments:

Post a Comment