With Elastic APM, you can capture system and process metrics. These metrics will be sent regularly to the APM Server and from there to Elasticsearch

Metric setsedit

CPU/Memory metric setedit


This metric set collects various system metrics and metrics of the current process.

if you do not use Linux, you need to install psutil for this metric set.

type: scaled_float

format: percent

The percentage of CPU time in states other than Idle and IOWait, normalized by the number of cores.

type: scaled_float

format: percent

The percentage of CPU time spent by the process since the last event. This value is normalized by the number of CPU cores and it ranges from 0 to 100%.

type: long

format: bytes

Total memory.

type: long

format: bytes

Actual free memory in bytes.


type: long

format: bytes

The total virtual memory the process has.


type: long

format: bytes

The Resident Set Size. The amount of memory the process occupied in main memory (RAM).

Linux’s cgroup metricsedit

type: long

format: bytes

Memory limit for current cgroup slice.


type: long

format: bytes

Memory usage in current cgroup slice.

Breakdown metric setedit

Tracking and collection of this metric set can be disabled using the breakdown_metrics setting.


type: simple timer

This timer tracks the span self-times and is the basis of the transaction breakdown visualization.


  • sum: The sum of all span self-times in ms since the last report (the delta)
  • count: The count of all span self-times since the last report (the delta)

You can filter and group by these dimensions:

  • The name of the transaction
  • transaction.type: The type of the transaction, for example request
  • span.type: The type of the span, for example app, template or db
  • span.subtype: The sub-type of the span, for example mysql (optional)

Prometheus metric set (beta)edit

This functionality is in beta and is subject to change. The design and code is less mature than official GA features and is being provided as-is with no warranties. Beta features are not subject to the support SLA of official GA features.

If you use prometheus_client to collect metrics, the agent can collect them as well and make them available in Elasticsearch.

The following types of metrics are supported:

  • Counters
  • Gauges
  • Summaries
  • Histograms (requires APM Server / Elasticsearch / Kibana 7.14+)

To use the Prometheus metric set, you have to enable it with the prometheus_metrics configuration option.

All metrics collected from prometheus_client are prefixed with "prometheus.metrics.". This can be changed using the prometheus_metrics_prefix configuration option.

Beta limitationsedit
  • The metrics format may change without backwards compatibility in future releases.

Custom Metricsedit

Custom metrics allow you to send your own metrics to Elasticsearch.

The most common way to send custom metrics is with the Prometheus metric set. However, you can also use your own metric set. If you collect the metrics manually in your code, you can use the base MetricSet class:

from elasticapm.metrics.base_metrics import MetricSet

client = elasticapm.Client()
metricset = client.metrics.register(MetricSet)

for x in range(10):

Alternatively, you can create your own MetricSet class which inherits from the base class. In this case, you’ll usually want to override the before_collect method, where you can gather and set metrics before they are collected and sent to Elasticsearch.

You can add your MetricSet class as shown in the example above, or you can add an import string for your class to the metrics_sets configuration option:


Your MetricSet might look something like this:

from elasticapm.metrics.base_metrics import MetricSet

class MyAwesomeMetricSet(MetricSet):
    def before_collect(self):

In the example above, the MetricSet would look up myapp.some_value and set the metric my_gauge to that value. This would happen whenever metrics are collected/sent, which is controlled by the metrics_interval setting.