This guide demonstrates how your Quarkus application can utilize the Micrometer metrics library for runtime and application metrics.
Apart from application-specific metrics, which are described in this guide, you may also utilize built-in metrics exposed by various Quarkus extensions. These are described in the guide for each particular extension that supports built-in metrics.
Important
|
Micrometer is the recommended approach to metrics for Quarkus. |
Micrometer defines a core library providing a registration mechanism for Metrics, and core metric types (Counters,
Gauges, Timers, Distribution Summaries, etc.). These core types provide an abstraction layer that can be adapted to
different backend monitoring systems. In essence, your application (or a library) can register
a Counter
,
Gauge
, Timer
, or DistributionSummary
with a MeterRegistry
. Micrometer will then delegate that registration to
one or more implementations, where each implementation handles the unique considerations for the associated
monitoring stack.
Micrometer uses naming conventions to translate between registered Meters and the conventions used by various backend
registries. Meter names, for example, should be created and named using dots to separate segments, a.name.like.this
.
Micrometer then translates that name into the format that the selected registry prefers. Prometheus
uses underscores, which means the previous name will appear as a_name_like_this
in Prometheus-formatted metrics
output.
We recommend that you follow the instructions in the next sections and create the application step by step. You can skip right to the solution if you prefer. Either:
-
Clone the git repository:
git clone {quickstarts-clone-url}
, or -
Download an {quickstarts-archive-url}[archive].
The solution is located in the micrometer-quickstart
{quickstarts-tree-url}/micrometer-quickstart[directory].
Quarkus Micrometer extensions are structured similarly to Micrometer itself: quarkus-micrometer
provides core
micrometer support and runtime integration and other Quarkus and Quarkiverse extensions bring in additional
dependencies and requirements to support specific monitoring systems.
For this example, we’ll use the Prometheus registry.
First, we need a new project. Create a new project with the following command:
This command generates a Maven project, that imports the micrometer-registry-prometheus
extension as a dependency.
This extension will load the core micrometer
extension as well as additional library dependencies required to support
prometheus.
If you already have your Quarkus project configured, you can add the micrometer-registry-prometheus
extension
to your project by running the following command in your project base directory:
This will add the following to your build file:
<dependency>
<groupId>io.quarkus</groupId>
<artifactId>quarkus-micrometer-registry-prometheus</artifactId>
</dependency>
implementation("io.quarkus:quarkus-micrometer-registry-prometheus")
Micrometer provides an API that allows you to construct your own custom metrics. The most common types of meters supported by monitoring systems are gauges, counters, and summaries. The following sections build an example endpoint, and observes endpoint behavior using these basic meter types.
To register meters, you need a reference to a MeterRegistry
, which is configured and maintained by the Micrometer
extension. The MeterRegistry
can be injected into your application as follows:
package org.acme.micrometer;
import io.micrometer.core.instrument.MeterRegistry;
import jakarta.ws.rs.GET;
import jakarta.ws.rs.Path;
import jakarta.ws.rs.Produces;
@Path("/example")
@Produces("text/plain")
public class ExampleResource {
private final MeterRegistry registry;
ExampleResource(MeterRegistry registry) {
this.registry = registry;
}
}
Micrometer maintains an internal mapping between unique metric identifier and tag combinations and specific meter
instances. Using register
, counter
, or other methods to increment counters or record values does not create
a new instance of a meter unless that combination of identifier and tag/label values hasn’t been seen before.
Gauges measure a value that can increase or decrease over time, like the speedometer on a car. Gauges can be useful when monitoring the statistics for a cache or collection. Consider the following simple example that observes the size of a list:
LinkedList<Long> list = new LinkedList<>();
// Update the constructor to create the gauge
ExampleResource(MeterRegistry registry) {
this.registry = registry;
registry.gaugeCollectionSize("example.list.size", Tags.empty(), list);
}
@GET
@Path("gauge/{number}")
public Long checkListSize(long number) {
if (number == 2 || number % 2 == 0) {
// add even numbers to the list
list.add(number);
} else {
// remove items from the list for odd numbers
try {
number = list.removeFirst();
} catch (NoSuchElementException nse) {
number = 0;
}
}
return number;
}
Note that even numbers are added to the list, and odd numbers remove an element from the list.
Start your application in dev mode:
Then try the following sequence and look for example_list_size
in the plain text output:
curl http://localhost:8080/example/gauge/1
curl http://localhost:8080/example/gauge/2
curl http://localhost:8080/example/gauge/4
curl http://localhost:8080/q/metrics
curl http://localhost:8080/example/gauge/6
curl http://localhost:8080/example/gauge/5
curl http://localhost:8080/example/gauge/7
curl http://localhost:8080/q/metrics
It is important to note that gauges are sampled rather than set; there is no record of how the value associated with a gauge might have changed between measurements. In this example, the size of the list is observed when the Prometheus endpoint is visited.
Micrometer provides a few additional mechanisms for creating gauges. Note that Micrometer does not create strong
references to the objects it observes by default. Depending on the registry, Micrometer either omits gauges that observe
objects that have been garbage-collected entirely or uses NaN
(not a number) as the observed value.
When should you use a Gauge? Only if you can’t use something else. Never gauge something you can count. Gauges can be less straight-forward to use than counters. If what you are measuring can be counted (because the value always increments), use a counter instead.
Counters are used to measure values that only increase. In the example below, you will count the number of times you test a number to see if it is prime:
@GET
@Path("prime/{number}")
public String checkIfPrime(long number) {
if (number < 1) {
return "Only natural numbers can be prime numbers.";
}
if (number == 1 || number == 2 || number % 2 == 0) {
return number + " is not prime.";
}
if ( testPrimeNumber(number) ) {
return number + " is prime.";
} else {
return number + " is not prime.";
}
}
protected boolean testPrimeNumber(long number) {
// Count the number of times we test for a prime number
registry.counter("example.prime.number").increment();
for (int i = 3; i < Math.floor(Math.sqrt(number)) + 1; i = i + 2) {
if (number % i == 0) {
return false;
}
}
return true;
}
It might be tempting to add a label or tag to the counter indicating what value was checked, but remember that each
unique combination of metric name (example.prime.number
) and label value produces a unique time series. Using an
unbounded set of data as label values can lead to a "cardinality explosion", an exponential increase in the creation
of new time series.
Note
|
Label and tag can be used interchangeably. You may also see "attribute" used in this context in some documentation.
The gist is each that each label or tag or attribute defines an additional bit of information associated with the
single numerical measurement that helps you classify, group, or aggregate the measured value later. The Micrometer API
uses |
It is possible to add a tag that would convey a little more information, however. Let’s adjust our code, and move the counter to add some tags to convey additional information.
@GET
@Path("prime/{number}")
public String checkIfPrime(long number) {
if (number < 1) {
registry.counter("example.prime.number", "type", "not-natural").increment();
return "Only natural numbers can be prime numbers.";
}
if (number == 1 ) {
registry.counter("example.prime.number", "type", "one").increment();
return number + " is not prime.";
}
if (number == 2 || number % 2 == 0) {
registry.counter("example.prime.number", "type", "even").increment();
return number + " is not prime.";
}
if ( testPrimeNumber(number) ) {
registry.counter("example.prime.number", "type", "prime").increment();
return number + " is prime.";
} else {
registry.counter("example.prime.number", "type", "not-prime").increment();
return number + " is not prime.";
}
}
protected boolean testPrimeNumber(long number) {
for (int i = 3; i < Math.floor(Math.sqrt(number)) + 1; i = i + 2) {
if (number % i == 0) {
return false;
}
}
return true;
}
Looking at the data produced by this counter, you can tell how often a negative number was checked, or the number one,
or an even number, and so on. Try the following sequence and look for example_prime_number_total
in the plain text
output. Note that the _total
suffix is added when Micrometer applies Prometheus naming conventions to
example.prime.number
, the originally specified counter name.
If you did not leave Quarkus running in dev mode, start it again:
Then execute the following sequence:
curl http://localhost:8080/example/prime/-1
curl http://localhost:8080/example/prime/0
curl http://localhost:8080/example/prime/1
curl http://localhost:8080/example/prime/2
curl http://localhost:8080/example/prime/3
curl http://localhost:8080/example/prime/15
curl http://localhost:8080/q/metrics
When should you use a counter? Only if you are doing something that can not be either timed (or summarized). Counters only record a count, which may be all that is needed. However, if you want to understand more about how a value is changing, a timer (when the base unit of measurement is time) or a distribution summary might be more appropriate.
Timers and distribution summaries in Micrometer are very similar. Both allow you to record an observed value, which will be aggregated with other recorded values and stored as a sum. Micrometer also increments a counter to indicate the number of measurements that have been recorded and tracks the maximum observed value (within a decaying interval).
Distribution summaries are populated by calling the record
method to record observed values, while timers provide
additional capabilities specific to working with time and measuring durations. For example, we can use a timer to
measure how long it takes to calculate prime numbers using one of the record
methods that wraps the invocation of a
Supplier function:
protected boolean testPrimeNumber(long number) {
Timer timer = registry.timer("example.prime.number.test");
return timer.record(() -> {
for (int i = 3; i < Math.floor(Math.sqrt(number)) + 1; i = i + 2) {
if (number % i == 0) {
return false;
}
}
return true;
});
}
Micrometer will apply Prometheus conventions when emitting metrics for this timer. Prometheus measures time in seconds.
Micrometer converts measured durations into seconds and includes the unit in the metric name, per convention. After
visiting the prime endpoint a few more times, look in the plain text output for the following three entries:
example_prime_number_test_seconds_count
, example_prime_number_test_seconds_sum
, and
example_prime_number_test_seconds_max
.
If you did not leave Quarkus running in dev mode, start it again:
Then execute the following sequence:
curl http://localhost:8080/example/prime/256
curl http://localhost:8080/q/metrics
curl http://localhost:8080/example/prime/7919
curl http://localhost:8080/q/metrics
Both timers and distribution summaries can be configured to emit additional statistics, like histogram data, precomputed percentiles, or service level objective (SLO) boundaries. Note that the count, sum, and histogram data can be re-aggregated across dimensions (or across a series of instances), while precomputed percentile values cannot.
To view metrics, execute curl localhost:8080/q/metrics/
The Micrometer extension automatically times HTTP server requests. Following Prometheus naming conventions for
timers, look for http_server_requests_seconds_count
, http_server_requests_seconds_sum
, and
http_server_requests_seconds_max
. Dimensional labels have been added for the requested uri, the HTTP method
(GET, POST, etc.), the status code (200, 302, 404, etc.), and a more general outcome field.
# HELP http_server_requests_seconds
# TYPE http_server_requests_seconds summary
http_server_requests_seconds_count{method="GET",outcome="SUCCESS",status="200",uri="/example/prime/{number}"} 1.0
http_server_requests_seconds_sum{method="GET",outcome="SUCCESS",status="200",uri="/example/prime/{number}"} 0.017385896
# HELP http_server_requests_seconds_max
# TYPE http_server_requests_seconds_max gauge
http_server_requests_seconds_max{method="GET",outcome="SUCCESS",status="200",uri="/example/prime/{number}"} 0.017385896
#
Note that metrics appear lazily, you often won’t see any data for your endpoint until something tries to access it, etc.
You can disable measurement of HTTP endpoints using the quarkus.micrometer.binder.http-server.ignore-patterns
property. This property accepts a comma-separated list of simple regex match patterns identifying URI paths that should
be ignored. For example, setting quarkus.micrometer.binder.http-server.ignore-patterns=/example/prime/[0-9]+
will
ignore a request to http://localhost:8080/example/prime/7919
. A request to http://localhost:8080/example/gauge/7919
would still be measured.
The micrometer extension will make a best effort at representing URIs containing path parameters in templated form.
Using examples from above, a request to http://localhost:8080/example/prime/7919
should appear as an attribute of
http_server_requests_seconds_*
metrics with a value of uri=/example/prime/{number}
.
Use the quarkus.micrometer.binder.http-server.match-patterns
property if the correct URL can not be determined. This
property accepts a comma-separated list defining an association between a simple regex match pattern and a replacement
string. For example, setting
quarkus.micrometer.binder.http-server.match-patterns=/example/prime/=/example/{jellybeans}` would use the value
`/example/{jellybeans}` for the uri attribute any time the requested uri matches `/example/prime/[0-9]
.
By default, the metrics are exported using the Prometheus format application/openmetrics-text
,
you can revert to the former format by specifying the Accept
request header to text/plain
(curl -H "Accept: text/plain" localhost:8080/q/metrics/
).
Micrometer uses MeterFilter
instances to customize the metrics emitted by MeterRegistry
instances.
The Micrometer extension will detect MeterFilter
CDI beans and use them when initializing MeterRegistry
instances.
@Singleton
public class CustomConfiguration {
@ConfigProperty(name = "deployment.env")
String deploymentEnv;
/** Define common tags that apply only to a Prometheus Registry */
@Produces
@Singleton
@MeterFilterConstraint(applyTo = PrometheusMeterRegistry.class)
public MeterFilter configurePrometheusRegistries() {
return MeterFilter.commonTags(Arrays.asList(
Tag.of("registry", "prometheus")));
}
/** Define common tags that apply globally */
@Produces
@Singleton
public MeterFilter configureAllRegistries() {
return MeterFilter.commonTags(Arrays.asList(
Tag.of("env", deploymentEnv)));
}
/** Enable histogram buckets for a specific timer */
@Produces
@Singleton
public MeterFilter enableHistogram() {
return new MeterFilter() {
@Override
public DistributionStatisticConfig configure(Meter.Id id, DistributionStatisticConfig config) {
if(id.getName().startsWith("myservice")) {
return DistributionStatisticConfig.builder()
.percentiles(0.5, 0.95) // median and 95th percentile, not aggregable
.percentilesHistogram(true) // histogram buckets (e.g. prometheus histogram_quantile)
.build()
.merge(config);
}
return config;
}
};
}
}
In this example, a singleton CDI bean will produce two different MeterFilter
beans. One will be applied only to
Prometheus MeterRegistry
instances (using the @MeterFilterConstraint
qualifier), and another will be applied
to all MeterRegistry
instances. An application configuration property is also injected and used as a tag value.
Additional examples of MeterFilters can be found in the
official documentation.
Micrometer does define two annotations, @Counted
and @Timed
, that can be added to methods. The @Timed
annotation
will wrap the execution of a method and will emit the following tags in addition to any tags defined on the
annotation itself: class, method, and exception (either "none" or the simple class name of a detected exception).
Using annotations is limited, as you can’t dynamically assign meaningful tag values. Also note that many methods, e.g. REST endpoint methods or Vert.x Routes, are counted and timed by the micrometer extension out of the box.
If you aren’t using Prometheus, you have a few options. Some Micrometer registry implementations
have been wrapped in
Quarkiverse extensions.
To use the Micrometer StackDriver MeterRegistry, for example, you would use the
quarkus-micrometer-registry-stackdriver
extension:
<dependency>
<groupId>io.quarkus</groupId>
<artifactId>quarkus-micrometer-registry-stackdriver</artifactId>
</dependency>
implementation("io.quarkus:quarkus-micrometer-registry-stackdriver")
If the Micrometer registry you would like to use does not yet have an associated extension,
use the quarkus-micrometer
extension and bring in the packaged MeterRegistry dependency directly:
<dependency>
<groupId>io.quarkus</groupId>
<artifactId>quarkus-micrometer</artifactId>
</dependency>
<dependency>
<groupId>com.acme</groupId>
<artifactId>custom-micrometer-registry</artifactId>
</dependency>
implementation("io.quarkus:quarkus-micrometer")
implementation("com.acme:custom-micrometer-registry")
You will then need to specify your own provider to configure and initialize the MeterRegistry, as discussed in the next section.
Use a custom @Produces
method to create and configure a customized MeterRegistry
if you need to.
The following example customizes the line format used for StatsD:
@Produces
@Singleton
public StatsdMeterRegistry createStatsdMeterRegistry(StatsdConfig statsdConfig, Clock clock) {
// define what to do with lines
Consumer<String> lineLogger = line -> logger.info(line);
// inject a configuration object, and then customize the line builder
return StatsdMeterRegistry.builder(statsdConfig)
.clock(clock)
.lineSink(lineLogger)
.build();
}
This example corresponds to the following instructions in the Micrometer documentation: https://micrometer.io/docs/registry/statsD#_customizing_the_metrics_sink
Note that the method returns the specific type of MeterRegistry
as a @Singleton
. Use MicroProfile Config
to inject any configuration attributes you need to configure the registry. Most Micrometer registry extensions,
like quarkus-micrometer-registry-statsd
, define a producer for registry-specific configuration objects
that are integrated with the Quarkus configuration model.
If you use the MicroProfile Metrics API in your application, the Micrometer extension will create an adaptive
layer to map those metrics into the Micrometer registry. Note that naming conventions between the two
systems is different, so the metrics that are emitted when using MP Metrics with Micrometer will change.
You can use a MeterFilter
to remap names or tags according to your conventions.
@Produces
@Singleton
public MeterFilter renameApplicationMeters() {
final String targetMetric = MPResourceClass.class.getName() + ".mpAnnotatedMethodName";
return MeterFilter() {
@Override
public Meter.Id map(Meter.Id id) {
if (id.getName().equals(targetMetric)) {
// Drop the scope tag (MP Registry type: application, vendor, base)
List<Tag> tags = id.getTags().stream().filter(x -> !"scope".equals(x.getKey()))
.collect(Collectors.toList());
// rename the metric
return id.withName("my.metric.name").replaceTags(tags);
}
return id;
}
};
}
Ensure the following dependency is present in your build file if you require the MicroProfile Metrics API:
<dependency>
<groupId>org.eclipse.microprofile.metrics</groupId>
<artifactId>microprofile-metrics-api</artifactId>
</dependency>
implementation("org.eclipse.microprofile.metrics:microprofile-metrics-api")
Note
|
The MP Metrics API compatibility layer will be moved to a different extension in the future. |