Spring Cloud Data Flow Metrics provides the foundation to help export Boot metrics that are exposed under the /metrics endpoint to a variety of external systems. It provides Spring Boot AutoConfiguration to setup the exporting process and common functionality such as defining a prefix for your metric names.

The spring-cloud-dataflow-metrics-log module provides a LogMetricWriter that writes Boot metrics into the log file. While it is a very simple MetricWriter implementation, log files are often ingested into application monitoring tools where they can be further processed to create dashboards of an application’s performance.

Most important for Spring Cloud Data Flow Stream applications are the inclusion of Spring Integration channel metrics such as message rates.

Add the Maven dependency

and reference the Spring Snapshot repository

The prefix that will be prepended to the names of metrics that are exposed under the /metrics endpoint can be configured using the property spring.cloud.dataflow.metrics.prefix.

The default value uses the form ${spring.cloud.application.group}.${spring.cloud.application.name}.<ID> where <ID> and the ${} placeholders are referencing standard Spring Boot Configuration properties.

If the spring.cloud.application placeholder values are not set, then the default value for ${spring.cloud.application.group} is group and ${spring.cloud.application.name} is application

The value of <ID> is resovled to one of the following placeholder expressions in the following order

• ${vcap.application.instance_index}

• ${spring.application.index}

• ${PID}

Other useful property placeholder keys that you should be aware of when setting your own prefix value are

• ${server.port}

• ${spring.cloud.client.ipAddress}

• ${spring.cloud.client.hostname}

The exporting process is controller though MetricExportProperties. This lets you control how often the metrics are exported and which metrics to include/exclude. Some documentation is available here.

The project Spring Cloud Data Flow Metrics Datadog exports metrics to Datadog.

Timemonitor is an example application (a WIP) that demonstrates creating the time source application from Data Flow App Starters and adding the ability to export metrics.

Planning a 1.0 milestone release end of November.