jaxrs-monitor

A tiny monitoring plugin for JavaEE applications. jaxrs-monitor collects and exposes metrics about the application, its runtime and response times. Metrics are exposed via a JAX-RS resource in JSON format.

This tool provides easy means to integrate your application into a centralized monitoring infrastructure.

Make sure that your application server provides JAX-RS, JSON and EJB implementations. jaxrs-monitor was tested on IBM WebSphere Liberty Profile (webProfile7).

This application is licensed under the BSD 2-clause licence.

Usage

Include the following in your Maven Pom File:

<dependency>
    <groupId>com.github.tnsasse</groupId>
    <artifactId>jaxrs-monitor</artifactId>
    <version>VERSION_GOES_HERE</version>
</dependency>

Substitute VERSION_GOES_HERE with the current version of your choice.

jaxrs-monitor provides a servlet filter to collect response times and will expose metrics via the path /jaxrs-monitor so make sure to avoid collisions.

You can configure the following parameters, by calling the MonitorConfiguration bean.

• Max monitoring depth: restrict the path depth that should be monitored for response times metrics.
• Enable/disable the app, system and response metrics individually.

Example Configuration:

    @Inject
    MonitoringConfiguration configuration

Set the desired settings, see the Javadoc for more information:

    configuration.setSystemMetricsEnabled(true);
    configuration.setAppMetricsEnabled(true);
    configuration.setResponseMetricsEnabled(true);
    configuration.setMaxPathDepth(-1);

Build / Install

You can clone this repository and build it with maven:

git clone https://github.com/tnsasse/jaxrs-monitor.git
cd jaxrs-monitor
mvn clean install

Exposed metrics

jaxrs-monitor exposes the following metrics, divided into categories:

System metrics

• Number of processors available to the jvm ("cores")
• Available memory in megabytes ("availableMemory")
• Used memory in megabytes ("usedMemory")
• Average system load ("loadAverage")

Application metrics

• Application startup time in ISO8601 ("startupTime")
• Uptime since application start in milliseconds ("uptime")

HTTP response times

Per HTTP request URI and HTTP method:

• Number of requests ("count")
• Minimum response time in milliseconds ("min")
• Maximum response time in milliseconds ("max")
• Average response time in milliseconds ("avg")

Sample application

Check out jaxrs-monitor-sample for a working demo of the application.

Sample output

All metrics:

GET http://URL:PORT/CONTEXT_ROOT/jaxrs-monitor
{
    "system" : {
        "cores" : 2,
        "availableMemory" : 26,
        "usedMemory" : 49,
        "loadAverage" : 0.65
    },
    "app" : { 
        "startupTime" : "2017-04-01T12:59:01.660Z[UTC]",
        "uptime" : 42648
    },
    "response" : {
        "/orders" : { 
            "GET" : {"count":11,"min":3,"max":18,"avg":7},
            "POST" : {"count":4,"min":23,"max":83,"avg":13}
        },
        "/profile" : {
            "GET" : {"count":14,"min":2,"max":15,"avg":8}
        }
    }
}

Only system metrics:

GET http://URL:PORT/CONTEXT_ROOT/jaxrs-monitor/system
{
    "cores" : 2,
    "availableMemory" : 26,
    "usedMemory" : 49,
    "loadAverage" : 0.65
}

Only app metrics:

GET http://URL:PORT/CONTEXT_ROOT/jaxrs-monitor/app
{
    "startupTime" : "2017-04-01T12:59:01.660Z[UTC]",
    "uptime" : 42648
}

Only response times:

GET http://URL_PORT/CONTEXT_ROOT/jaxrs-monitor/response
{
    "/orders" : { 
        "GET" : {"count":11,"min":3,"max":18,"avg":7},
        "POST" : {"count":4,"min":23,"max":83,"avg":13}
    },
    "/profile" : {
            "GET" : {"count":14,"min":2,"max":15,"avg":8}
    }
}