porcupine

Configurable threading, resiliency and monitoring with injectable statistics for Java EE 7. Porcupine is the implementation of the bulkhead and handshaking patterns for Java EE 7.

Features:

Conventional: ExecutorService is directly injectable. The thread pool derives the name from the field, but can be easily overridden.
Drop-in install: a single dependency in the pom.xml is sufficient.
Standard based: porcupine uses JSR 236: Concurrency Utilities for JavaTM EE
Small: the entire framework is: 16kB.
Extensible without configuration: All major components can be replaced (big thanks to @Specializes)

Booting Porcupine

Configuration

Dealing with Overload

Automatic Statistics Injection

##Installation

        <dependency>
            <groupId>com.airhacks</groupId>
            <artifactId>porcupine</artifactId>
            <version>NEWEST_VERSION</version>
            <scope>compile</scope>
        </dependency>

With statistics injection into HTTP headers of all JAX-RS resources:

        <dependency>
            <groupId>com.airhacks</groupId>
            <artifactId>porcupine-spy</artifactId>
            <version>NEWEST_VERSION</version>
            <scope>compile</scope>
        </dependency>

##Conventional Usage

@Stateless
public class MessagesService {

    @Inject
    @Dedicated
    ExecutorService light;

    @Inject
    @Dedicated
    ExecutorService heavy;

Custom naming is also supported:

    @Inject
    @Dedicated("custom-pool")
    private ExecutorService first;

##Statistics and monitoring

###Exposure

@Path("statistics")
@RequestScoped
@Produces({MediaType.APPLICATION_XML, MediaType.APPLICATION_JSON})
public class StatisticsResource {

    @Inject
    Instance<List<Statistics>> statistics;

	@GET
    public List<Statistics> expose() {
        return this.statistics.get();
	}

}

###Sample output

Unified thread pool statistics are accessible via injection. Individual statistics are injectable with the @Dedicated qualifier.

XML statistics

<statistics>
	<pipelineName>light</pipelineName>
	<remainingQueueCapacity>100</remainingQueueCapacity>
	<completedTaskCount>1</completedTaskCount>
	<activeThreadCount>0</activeThreadCount>
	<largestThreadPoolSize>1</largestThreadPoolSize>
	<currentThreadPoolSize>1</currentThreadPoolSize>
	<totalNumberOfTasks>1</totalNumberOfTasks>
	<maximumPoolSize>48</maximumPoolSize>
	<rejectedExecutionHandlerName>ExecutorServiceExposer$$Lambda$3/1562863014</rejectedExecutionHandlerName>
	<rejectedTasks>0</rejectedTasks>
</statistics>
<statistics>
	<pipelineName>heavy</pipelineName>
	<remainingQueueCapacity>16</remainingQueueCapacity>
	<completedTaskCount>0</completedTaskCount>
	<activeThreadCount>1</activeThreadCount>
	<largestThreadPoolSize>1</largestThreadPoolSize>
	<currentThreadPoolSize>1</currentThreadPoolSize>
	<totalNumberOfTasks>1</totalNumberOfTasks>
	<maximumPoolSize>8</maximumPoolSize>
	<rejectedExecutionHandlerName>CallerRunsPolicy</rejectedExecutionHandlerName>
	<rejectedTasks>0</rejectedTasks>
</statistics>

JSON statistics

[{
        "pipelineName": "light",
        "remainingQueueCapacity": 100,
        "completedTaskCount": 1,
        "activeThreadCount": 0,
        "largestThreadPoolSize": 1,
        "currentThreadPoolSize": 1,
        "totalNumberOfTasks": 1,
        "maximumPoolSize": 48,
        "rejectedExecutionHandlerName": "ExecutorServiceExposer$$Lambda$3/1562863014",
        "rejectedTasks": 0
    },
    {
        "pipelineName": "heavy",
        "remainingQueueCapacity": 16,
        "completedTaskCount": 1,
        "activeThreadCount": 0,
        "largestThreadPoolSize": 1,
        "currentThreadPoolSize": 1,
        "totalNumberOfTasks": 1,
        "maximumPoolSize": 8,
        "rejectedExecutionHandlerName": "CallerRunsPolicy",
        "rejectedTasks": 0
    }]

Custom configuration (fully optional)

@Specializes
public class CustomExecutorConfigurator extends ExecutorConfigurator {

    @Override
    public ExecutorConfiguration defaultConfigurator() {
        return super.defaultConfigurator();
    }

    @Override
    public ExecutorConfiguration forPipeline(String name) {
        if ("heavy".equals(name)) {
            return new ExecutorConfiguration.Builder().
                    corePoolSize(4).
                    maxPoolSize(8).
                    queueCapacity(16).
                    keepAliveTime(1).
                    callerRunsPolicy().
                    build();
        }
        return super.forPipeline(name);
    }

}

Statistics Injection into HTTP Headers

HTTP/1.1 200 OK
Server: GlassFish Server Open Source Edition  4.1 
X-Powered-By: Servlet/3.1 JSP/2.3 (GlassFish Server Open Source Edition  4.1  Java/Oracle Corporation/1.8)

x-porcupine-statistics-light: {"pipelineName":"light","activeThreadCount":1,"completedTaskCount":1,
"corePoolSize":8,"currentThreadPoolSize":2,"largestThreadPoolSize":2,"maximumPoolSize":16,"rejectedTasks":0,
"remainingQueueCapacity":100,"minQueueCapacity":100,"totalNumberOfTasks":2}
x-porcupine-statistics-heavy: {"pipelineName":"heavy","rejectedExecutionHandlerName":"CallerRunsPolicy",
"activeThreadCount":0,"completedTaskCount":1,"corePoolSize":4,"currentThreadPoolSize":1,
"largestThreadPoolSize":1,"maximumPoolSize":8,"rejectedTasks":0,"remainingQueueCapacity":16,
"minQueueCapacity":16,"totalNumberOfTasks":1}

Content-Type: text/plain
Date: Wed, 18 Mar 2015 07:50:20 GMT
Content-Length: 36

cannot override porcupine thread factory by @Specializes

Hi Adam,

I am not able to override porcupine default configuration. I wrote this class:


import com.airhacks.porcupine.configuration.control.ExecutorConfigurator;
import com.airhacks.porcupine.execution.control.ExecutorConfiguration;

import javax.enterprise.inject.Specializes;

@Specializes
public class CustomExecutorConfigurator extends ExecutorConfigurator {

    @Override
    public ExecutorConfiguration defaultConfigurator() {
        return super.defaultConfigurator();
    }

    @Override
    public ExecutorConfiguration forPipeline(String name) {
        return new ExecutorConfiguration.Builder().
                corePoolSize(4).
                maxPoolSize(83).
                queueCapacity(16).
                keepAliveTime(1).
                callerRunsPolicy().
                build();
    }

}

but when I hit statistics endpoint a I see still default configuration.

 {
        "pipelineName": "archiveBrowsableDays-endpoint-executor",
        "remainingQueueCapacity": 100,
        "minQueueCapacity": 100,
        "completedTaskCount": 0,
        "activeThreadCount": 0,
        "largestThreadPoolSize": 0,
        "currentThreadPoolSize": 0,
        "totalNumberOfTasks": 0,
        "maximumPoolSize": 16,
        "rejectedExecutionHandlerName": "ExecutorServiceExposer$$Lambda$3957/1578815540",
        "rejectedTasks": 0,
        "corePoolSize": 8
    },
    {
        "pipelineName": "init-endpoint-executor",
        "remainingQueueCapacity": 100,
        "minQueueCapacity": 100,
        "completedTaskCount": 1,
        "activeThreadCount": 0,
        "largestThreadPoolSize": 1,
        "currentThreadPoolSize": 1,
        "totalNumberOfTasks": 1,
        "maximumPoolSize": 16,
        "rejectedExecutionHandlerName": "ExecutorServiceExposer$$Lambda$3957/1578815540",
        "rejectedTasks": 0,
        "corePoolSize": 8
    }

What I do wrong? Thanks.

adambien / porcupine Goto Github PK

porcupine's Introduction

porcupine

Booting Porcupine

Configuration

Dealing with Overload

Automatic Statistics Injection

Custom configuration (fully optional)

Statistics Injection into HTTP Headers

porcupine's People

Contributors

Stargazers

Watchers

Forkers

porcupine's Issues

Recommend Projects

Recommend Topics

Recommend Org