Documentation Beyla Configure Options

Beyla configuration options

Beyla can be configured via environment variables or via a YAML configuration file that is passed either with the -config command-line argument or the BEYLA_CONFIG_PATH environment variable. Environment variables have priority over the properties in the configuration file. For example, in the following command line, the BEYLA_OPEN_PORT option, is used to override any open_port settings inside the config.yaml file:

Copy

$ BEYLA_OPEN_PORT=8080 beyla -config /path/to/config.yaml

or

Copy

$ BEYLA_OPEN_PORT=8080 BEYLA_CONFIG_PATH=/path/to/config.yaml beyla

At the end of this document, there is an example of YAML configuration file.

Currently, Beyla consist of a pipeline of components which generate, transform, and export traces from HTTP and GRPC applications. In the YAML configuration, each component has its own first-level section.

Optionally, Beyla also provides network-level metrics, which are documented in the Network metrics section of the Beyla documentation.

A quick description of the components:

• Service discovery searches for instrumentable processes matching a given criteria.
• EBPF tracer instruments the HTTP and GRPC services of an external process, creates service traces and forwards them to the next stage of the pipeline.
• Configuration of metrics and traces attributes to control which attributes are reported.
• Routes decorator will match HTTP paths (e.g. /user/1234/info) into user-provided HTTP routes (e.g. /user/{id}/info). If no routes are defined, the incoming data will be directly forwarded to the next stage.
• Kubernetes decorator will decorate the metrics and traces with Kubernetes metadata of the instrumented Pods.
• Filter metrics and traces by attribute values.
• Grafana Cloud OTEL exporter for metrics and traces simplifies the submission of OpenTelemetry metrics and traces to Grafana cloud.
• OTEL metrics exporter exports metrics data to an external OpenTelemetry metrics collector.
• OTEL traces exporter exports span data to an external OpenTelemetry traces collector.
• Prometheus HTTP endpoint enables an HTTP endpoint that allows any external scraper to pull metrics in Prometheus format.
• Internal metrics reporter optionally reports metrics about the internal behavior of the auto-instrumentation tool in Prometheus format.

The following sections explain the global configuration properties, as well as the options for each component.

Global configuration properties

The properties in this section are first-level YAML properties, as they apply to the whole Beyla configuration:

Selects the process to instrument by the executable name path. This property accepts a regular expression to be matched against the full executable command line, including the directory where the executable resides on the file system.

This property is used to select a single process to instrument, or a group of processes of similar characteristics. For more fine-grained process selection and grouping, you can follow the instructions in the service discovery section.

If the open_port property is set, the executable to be selected needs to match both properties.

When instrumenting by using the executable name, choose a non-ambiguous name, a name that will match a single executable on the target system. For example, if you set BEYLA_EXECUTABLE_NAME=server, and you have running two processes whose executables have the following paths:

sh Copy

/usr/local/bin/language-server
/opt/app/server

Beyla will match indistinctly one of the above processes and instrument both. If you just want to instrument one of them, you should be as concrete as possible about the value of the setting. For example, BEYLA_EXECUTABLE_NAME=/opt/app/server or just BEYLA_EXECUTABLE_NAME=/server.

Selects the process to instrument by the port it has open (listens to). This property accepts a comma-separated list of ports (for example, 80), and port ranges (for example, 8000-8999). If the executable matching only one of the ports in the list, it is considered to match the selection criteria.

For example, specifying the following property:

Copy

open_port: 80,443,8000-8999

Would make Beyla to select any executable that opens port 80, 443, or any of the ports between 8000 and 8999 included.

This property is used to select a single process to instrument, or a group of processes of similar characteristics. For more fine-grained process selection and grouping, you can follow the instructions in the service discovery section.

If the executable_name property is set, the executable to be selected needs to match both properties.

If an executable opens multiple ports, only one of the ports needs to be specified for Beyla to instrument all the HTTP/S and GRPC requests on all application ports. At the moment, there is no way to restrict the instrumentation only to the methods exposed through a specific port.

If the specified port range is wide (e.g. 1-65535) Beyla will try to execute all the processes owning one of the ports in the range.

Overrides the name of the instrumented service to be reported by the metrics exporter. Defining this property is equivalent to add a name entry into the discovery.services YAML section.

If a single instance of Beyla is instrumenting multiple instances of different processes, they will share the same service name even if they are different. If you need that a single instance of Beyla report different service names, follow the instructions in the service discovery section.

Optionally, allows assigning a namespace for the service selected from the executable_name or open_port properties.

Defining this property is equivalent to add a name entry into the discovery.services YAML section.

This will assume a single namespace for all the services instrumented by Beyla. If you need that a single instance of Beyla groups multiple services into different namespaces, follow the instructions in the service discovery section.

It is important to notice that this namespace is not a selector for Kubernetes namespaces. Its value will be use to set the value of standard telemetry attributes. For example, the OpenTelemetry service.namespace attribute.

Sets the verbosity level of the process standard output logger. Valid log level values are: DEBUG, INFO, WARN and ERROR. DEBUG being the most verbose and ERROR the least verbose.

Prints any instrumented trace on the standard output. The value of this option specify the format to be used when printing the trace. Valid formats are:

If you have set the enforce_sys_caps to true, if the required system capabilities are not present Beyla aborts its startup and logs a list of the missing capabilities.

If you have set the configuration option to false, Beyla logs a list of the missing capabilities only.

Service discovery

The executable_name, open_port, service_name and service_namespace are top-level properties that simplify the configuration of Beyla to instrument a single service, or a group of related services.

In some scenarios, Beyla will instrument a big variety of services; for example, as a Kubernetes DaemonSet that instruments all the services in a node. The discovery YAML section will let you specify a higher differentiation degree in the services that Beyla can instrument.

For example, it will allow overriding the service name and namespace per service type.

This section allows specifying different selection criteria for different services, as well as overriding some of their metadata, such as their reported name or namespace.

For more details about this section, go to the discovery services section of this document.

This section allows for specifying selection criteria for excluding services from being instrumented. It follows the same definition format as described in the discovery services section of this document.

This option is useful for avoiding instrumentation of services which are typically found in observability environments. For example, use this option to exclude instrumenting Prometheus, the OpenTelemetry collector or Grafana Alloy.

Disables the detection of Go specifics when the ebpf tracer inspects executables to be instrumented. The tracer will fallback to using generic instrumentation, which will generally be less efficient.

Disables Beyla instrumentation of services which are already instrumented with OpenTelemetry. Since Beyla is often deployed to monitor all services in a Kubernetes cluster, monitoring already instrumented services can lead to duplicate telemetry data, unless the instrumentation selection (or exclusion) criteria is carefully crafted. To avoid unnecessary configuration overhead, Beyla monitors for the OpenTelemetry SDK calls to publish metrics and traces, and automatically turns off instrumentation of services which publish their own telemetry data. Turn this option off if your application generated telemetry data doesn’t conflict with the Beyla generated metrics and traces.

Discovery services section

Example of YAML file allowing the selection of multiple groups of services:

yaml Copy

discovery:
  services:
    - exe_path: (worker)|(backend)|(frontend)
      namespace: MyApplication
    - exe_path: loadgen
      namespace: testing
      name: "TestLoadGenerator"

The above example YAML will select two groups of executables. The first group will be formed by any process whose executable path contains the worker, backend or frontend text. For each service, Beyla will take the service name attribute from the executable name. The reported service namespace for all the processes matching this group will be MyApplication.

The second group in the above example YAML will select any executable whose path contains regexp, but instead of taking the service name from the executable name, it will override the service name with TestLoadGenerator.

The rest of this section describes the properties that are accepted in each entry of the services list.

Each services entry is a map where the properties can be grouped according to two purposes:

• Overriding the reported service name and namespace: name and namespace properties.
• Selecting the process to instrument: the rest of the properties, referred as selectors in this documentation.

Defines a name for the matching instrumented service. It will be used to populate the service.name OTEL property and the service_name Prometheus property in the exported metrics/traces.

If the property is not set, it will default to any of the following properties, in order of precedence:

• If Kubernetes is enabled:
  1. The name of the Deployment that runs the instrumented process, if any.
  2. The name of the ReplicaSet/DaemonSet/StatefulSet that runs the instrumented process, if any.
  3. The name of the Pod that runs the instrumented process.
• If Kubernetes is not enabled:
  1. The name of the process executable file.

If multiple processes match the service selection criteria described below, the metrics and traces for all the instances might share the same service name; for example, when multiple instrumented processes run under the same Deployment, or have the same executable name. In that case, the reported instance attribute would allow differentiating the different instances of the service.

Defines a namespace for the matching instrumented service. If the property is not set, it will be defaulted to the Kubernetes namespace of that runs the instrumented process, if Kubernetes is available, or empty when Kubernetes is not available.

It is important to notice that this namespace is not a selector for Kubernetes namespaces. Its value will be use to set the value of standard telemetry attributes. For example, the OpenTelemetry service.namespace attribute.

Selects the process to instrument by the port it has open (listens to). This property accepts a comma-separated list of ports (for example, 80), and port ranges (for example, 8000-8999). If the executable matching only one of the ports in the list, it is considered to match the selection criteria.

For example, specifying the following property:

Copy

open_port: 80,443,8000-8999

Would make Beyla to select any executable that opens port 80, 443, or any of the ports between 8000 and 8999 included.

If other selectors are specified in the same services entry, the processes to be selected need to match all the selector properties.

If an executable opens multiple ports, only one of the ports needs to be specified for Beyla to instrument all the HTTP/S and GRPC requests on all application ports. At the moment, there is no way to restrict the instrumentation only to the methods exposed through a specific port.

Selects the processes to instrument by their executable name path. This property accepts a regular expression to be matched against the full executable command line, including the directory where the executable resides on the file system.

Beyla will try to instrument all the processes with an executable path matching this property. For example, setting exe_path: .* will make Beyla to try to instrument all the executables in the host.

If other selectors are specified in the same services entry, the processes to be selected need to match all the selector properties.

This selector property will limit the instrumentation to the applications running in the Kubernetes Namespaces with a name matching the provided regular expression.

If other selectors are specified in the same services entry, the processes to be selected need to match all the selector properties.

This selector property will limit the instrumentation to the applications running in the Kubernetes Pods with a name matching the provided regular expression.

If other selectors are specified in the same services entry, the processes to be selected need to match all the selector properties.

This selector property will limit the instrumentation to the applications running in the Kubernetes Deployments with a name matching the provided regular expression.

If other selectors are specified in the same services entry, the processes to be selected need to match all the selector properties.

This selector property will limit the instrumentation to the applications running in the Kubernetes ReplicaSets with a name matching the provided regular expression.

If other selectors are specified in the same services entry, the processes to be selected need to match all the selector properties.

This selector property will limit the instrumentation to the applications running in the Kubernetes StatefulSets with a name matching the provided regular expression.

If other selectors are specified in the same services entry, the processes to be selected need to match all the selector properties.

This selector property will limit the instrumentation to the applications running in the Kubernetes DaemonSet with a name matching the provided regular expression.

If other selectors are specified in the same services entry, the processes to be selected need to match all the selector properties.

This selector property will limit the instrumentation to the applications running in the Pods having owned by either a Deployment, ReplicaSet, DaemonSet or StatefulSet with a name matching the provided regular expression.

If other selectors are specified in the same services entry, the processes to be selected need to match all the selector properties.

This selector property limits the instrumentation to the applications running in the Pods having labels with keys matching the provided value as regular expression.

If other selectors are specified in the same services entry, the processes to be selected need to match all the selector properties.

For example:

yaml Copy

discovery:
  services:
    - k8s_namespace: frontend
      k8s_pod_labels:
        instrument: beyla

The preceding example discovers all Pods in the frontend namespace that have a label instrument with a value that matches the regular expression beyla.

EBPF tracer

YAML section ebpf.

Specifies how many messages need to be accumulated in the eBPF ringbuffer before sending a wake-up request to the user space code.

In high-load services (in terms of requests/second), tuning this option to higher values can help with reducing the CPU overhead of Beyla.

In low-load services (in terms of requests/second), high values of wakeup_len could add a noticeable delay in the time the metrics are submitted and become externally visible.

Enables tracking of request headers for the purposes of processing any incoming ‘Traceparent’ header values. If this option is enabled, when Beyla encounters an incoming server request with a ‘Traceparent’ header value, it will use the provided ’trace id’ to create its own trace spans.

This option does not have an effect on Go applications, where the ‘Traceparent’ field is always processed, without additional tracking of the request headers.

Enabling this option may increase the performance overhead in high request volume scenarios. This option is only useful when generating Beyla traces, it does not affect generation of Beyla metrics.

Configures the time interval after which an HTTP request is considered as a timeout. This option allows Beyla to report HTTP transactions which timeout and never return. To disable the automatic HTTP request timeout feature, set this option to zero, i.e. “0ms”.

Configures the HTTP tracer heuristic to send telemetry events as soon as a response is detected. Setting this option reduces the accuracy of timings for requests with large responses, however, in high request volume scenarios this option will reduce the number of dropped trace events.

Configuration of metrics and traces attributes

Grafana Beyla allows configuring how some attributes for metrics and traces are decorated. Under the attributes top YAML sections, you can enable other subsections configure how some attributes are set.

Selection of metric attributes

The Beyla exported metrics document lists the attributes that can be reported with each metric. Some of the attributes are reported by default while others are hidden to control the cardinality.

For each metric, you can control which attributes to see with the select subsection, which is a map where each key is the name of a metric (either in its OpenTelemetry or Prometheus port), and each metric has two more sub-properties: include and exclude.

• include is a list of attributes that need to be reported. Each attribute can be an attribute name or a wildcard (for example, k8s.dst.* to include all the attributes starting with k8s.dst). If no include list is provided, the default attribute set is reported (check Beyla exported metrics for more information about the default attributes for a given metric).
• exclude is a list to of attribute names/wildcards containing the attributes to remove from the include list (or the default attribute set).

Example:

yaml Copy

attributes:
  select:
    beyla_network_flow_bytes:
      # limit the beyla_network_flow_bytes attributes to only the three attributes
      include:
        - beyla.ip
        - src.name
        - dst.port
    sql_client_duration:
      # report all the possible attributes but db_statement
      include: ["*"]
      exclude: ["db_statement"]
    http_client_request_duration:
      # report the default attribute set but exclude the Kubernetes Pod information
      exclude: ["k8s.pod.*"]

Additionally, you can use “*” wildcards as metric names to add and exclude attributes for groups of metrics having the same name. For example:

yaml Copy

attributes:
  select:
    http_*:
      include: ["*"]
      exclude: ["http_path", "http_route"]
    http_client_*:
      # override http_* exclusion
      include: ["http_path"]
    http_server_*:
      # override http_* exclusion
      include: ["http_route"]

In the previous example, all the metrics with a name starting with http_ (or http.) would include all the possible attributes but http_path and http_route (or http.path/http.route). The http_client_* and http_server_* sections would override the base configuration, enabling the http_path attribute for the HTTP client metrics and http_route for the HTTP server metrics.

When a metric name matches multiple definitions using wildcards, exact matches have higher precedence than wild card matches.

Instance ID decoration

The metrics and the traces are decorated with a unique instance ID string, identifying each instrumented application. By default, Beyla uses the host name that runs Beyla (can be a container or Pod name), followed by the PID of the instrumented process; but you can override how the instance ID is composed in the instance_id YAML subsection under the attributes top-level section.

For example:

yaml Copy

attributes:
  instance_id:
    dns: false

If true, it will try to resolve the Beyla local hostname against the network DNS. If false, it will use the local hostname.

If set, the host part of the Instance ID will use the provided string instead of trying to automatically resolve the host name.

This option takes precedence over dns.

Kubernetes decorator

If you run Beyla in a Kubernetes environment, you can configure it to decorate the traces and metrics with the Standard OpenTelemetry labels:

• k8s.namespace.name
• k8s.deployment.name
• k8s.statefulset.name
• k8s.replicaset.name
• k8s.daemonset.name
• k8s.node.name
• k8s.pod.name
• k8s.pod.uid
• k8s.pod.start_time
• k8s.cluster.name

In YAML, this section is named kubernetes, and is located under the attributes top-level section. For example:

yaml Copy

attributes:
  kubernetes:
    enable: true

It is IMPORTANT to consider that enabling this feature requires a previous step of providing some extra permissions to the Beyla Pod. Consult the “Configuring Kubernetes metadata decoration section” in the “Running Beyla in Kubernetes” page.

If set to true, Beyla will decorate the metrics and traces with Kubernetes metadata.

If set to false, the Kubernetes metadata decorator will be disabled.

If set to autodetect, Beyla will try to automatically detect if it is running inside Kubernetes, and enable the metadata decoration if that is the case.

This is a standard Kubernetes configuration environment variable, and is used to tell Beyla where to find the Kubernetes configuration in order to try to establish communication with the Kubernetes Cluster.

Usually you won’t need to change this value.

The accepted value is a list that might contain node and service.

This option allows you to selectively disable some Kubernetes informers, which are continuously listening to the Kubernetes API to obtain the metadata that is required for decorating network metrics or application metrics and traces.

When Beyla is deployed as a DaemonSet in very large clusters, all the Beyla instances creating multiple informers might end up overloading the Kubernetes API.

Disabling some informers would cause reported metadata to be incomplete, but reduces the load of the Kubernetes API.

The Pods informer can’t be disabled. For that purpose, you should disable the whole Kubernetes metadata decoration.

Maximum time that Beyla waits for getting all the Kubernetes metadata before starting to decorate metrics and traces. If this timeout is reached, Beyla starts normally but the metadata attributes might be incomplete until all the Kubernetes metadata is locally updated in background.

Beyla is subscribed to immediately receive any update on resources’ metadata. In addition, Beyla periodically resynchronizes the whole Kubernetes metadata at the frequency specified by this property.

Higher values reduce the load on the Kubernetes API service.

Routes decorator

YAML section routes.

This section can be only configured via the YAML file. If no routes section is provided in the YAML file, a default routes’ pipeline stage will be created and filtered with the wildcard routes decorator.

Will match the provided URL path patterns and set the http.route trace/metric property accordingly. You should use the routes property whenever possible to reduce the cardinality of generated metrics.

Each route pattern is a URL path with specific tags which allow for grouping path segments. The matcher tags can be in the :name or {name} format.

For example, if you define the following patterns:

yaml Copy

routes:
  patterns:
    - /user/{id}
    - /user/{id}/basket/{product}

Traces with the following HTTP paths will include the same http.route='/user/{id}' property:

Copy

/user/123
/user/456

Traces with the following HTTP paths will include the same http.route='/user/{id}'/basket/{product} property:

Copy

/user/123/basket/1
/user/456/basket/3

Additionally, the route matcher also supports the wildcard character *, which can be used to match path prefixes. For example, if you define the following pattern:

yaml Copy

routes:
  patterns:
    - /user/*

Any traces with HTTP paths starting with /user (including /user itself) will be matched to the route /user/*. As per the example above, all of the following paths will be matched as /user/*:

Copy

/user
/user/123
/user/123/basket/1
/user/456/basket/3

Will match the provided URL path against the defined patterns, and discard the trace and/or metric events if they match any of the ignored_patterns. The format for the ignored_patterns field is identical to the patterns field described above. You can define the ignored patterns with or without any of the wildcard options. For example, if you define the following ignored patterns:

yaml Copy

routes:
  ignored_patterns:
    - /health
    - /v1/*

Any event paths which have a prefix of /v1 or are equal to /health will be ignored.

This option is very useful if you want to prevent certain paths used development or service health monitoring, to be recorded as traces or metrics.

This property can be used together with the ignored_patterns property to refine which type of events are ignored.

Possible values for the ignore_mode property are:

• all will discard both metrics and traces which match the ignored_patterns.
• traces will discard only the traces which match the ignored_patterns. No metric events will be ignored.
• metrics will discard only the metrics which match the ignored_patterns. No trace events will be ignored.

Selectively ignoring only certain type of events might be useful in certain scenarios. For example, you may want to know the performance metrics of your health check API, but you wouldn’t want the overhead of those trace records in your target traces database. In this example scenario, you would set the ignore_mode property to traces, such that only traces matching the ignored_patterns will be discarded, while metrics will still be recorded.

Specifies what to do when a trace HTTP path does not match any of the patterns entries.

Possible values for the unmatched property are:

• unset will leave the http.route property as unset.
• path will copy the http.route field property to the path value.
  • 🚨 Caution: this option could lead to cardinality explosion at the ingester side.
• wildcard will set the http.route field property to a generic asterisk based /** value.
• heuristic will automatically derive the http.route field property from the path value, based on the following rules:
  • Any path components which have numbers or characters outside of the ASCII alphabet (or - and _), will be replaced by an asterisk *.
  • Any alphabetical components which don’t look like words, will be replaced by an asterisk *.

Special considerations when using the heuristic route decorator mode

The heuristic decorator is a best effort route decorator, which may still lead to cardinality explosion in certain scenarios. For example, the GitHub URL paths are a good example where the heuristic route decorator will not work, since the URL paths are constructed like a directory tree. In this scenario all paths will remain unique and lead to cardinality explosion.

On the other hand, if your URL path patterns follow certain structure, and the unique IDs are made up of numbers or random characters, then the heuristic decorator may be a low effort configuration option which is suitable for your use-case. For example, the following mock Google Docs URLs will be correctly reduced to a low cardinality version:

Both URL paths below:

Copy

document/d/CfMkAGbE_aivhFydEpaRafPuGWbmHfG/edit (no numbers in the ID)
document/d/C2fMkAGb3E_aivhFyd5EpaRafP123uGWbmHfG/edit

will be converted to a low cardinality route:

Copy

document/d/*/edit

OTEL metrics exporter

ℹ️ If you plan to use Beyla to send metrics to Grafana Cloud, consult the Grafana Cloud OTEL exporter for metrics and traces section for easier configuration.

YAML section otel_metrics_export.

This component exports OpenTelemetry metrics to a given endpoint. It will be enabled if its endpoint attribute is set (either via an YAML configuration file or via environment variables).

In addition to the properties exposed in this section, this component implicitly supports the environment variables from the standard OTEL exporter configuration.

Specifies the OpenTelemetry endpoint where metrics will be sent. If you plan to send the metrics directly to the Grafana Cloud OpenTelemetry endpoint, you might prefer to use the configuration options in the Using the Grafana Cloud OTEL endpoint to ingest metrics and traces section.

The OTEL_EXPORTER_OTLP_ENDPOINT environment variable sets a common endpoint for both the metrics and the traces exporters. The OTEL_EXPORTER_OTLP_METRICS_ENDPOINT environment variable, or the endpoint YAML, property will set the endpoint only for the metrics exporter node, such that the traces’ exporter won’t be activated unless explicitly specified.

According to the OpenTelemetry standard, if you set the endpoint via the OTEL_EXPORTER_OTLP_ENDPOINT environment variable, the OpenTelemetry exporter will automatically add the /v1/metrics path to the URL. If you want to avoid this addition, you can use either the OTEL_EXPORTER_OTLP_METRICS_ENDPOINT environment variable or the environment YAML property to use exactly the provided URL without any addition.

Specifies the transport/encoding protocol of the OpenTelemetry endpoint.

The accepted values, as defined by the OTLP Exporter Configuration document are http/json, http/protobuf and grpc.

The OTEL_EXPORTER_OTLP_PROTOCOL environment variable sets a common protocol for both the metrics and traces exporters. The OTEL_EXPORTER_OTLP_METRICS_PROTOCOL environment variable, or the protocol YAML property, will set the protocol only for the metrics exporter node.

If this property is not provided, Beyla will guess it according to the following rules:

• Beyla will guess grpc if the port ends in 4317 (4317, 14317, 24317, …), as 4317 is the usual Port number for the OTEL GRPC collector.
• Beyla will guess http/protobuf if the port ends in 4318 (4318, 14318, 24318, …), as 4318 is the usual Port number for the OTEL HTTP collector.

Controls whether the OTEL client verifies the server’s certificate chain and host name. If set to true, the OTEL client accepts any certificate presented by the server and any host name in that certificate. In this mode, TLS is susceptible to a man-in-the-middle attacks. This option should be used only for testing and development purposes.

Configures the intervening time between exports.

A list of metric groups which are allowed to be exported. Each group belongs to a different feature of Beyla: application-level metrics or network metrics.

• If the list contains application, the Beyla OpenTelemetry exporter exports application-level metrics; but only if there is defined an OpenTelemetry endpoint, and Beyla was able to discover any process matching the entries in the discovery section.
• If the list contains application_span, the Beyla OpenTelemetry exporter exports application-level trace span metrics; but only if there is defined an OpenTelemetry endpoint, and Beyla was able to discover any process matching the entries in the discovery section.
• If the list contains application_service_graph, the Beyla OpenTelemetry exporter exports application-level service graph metrics; but only if there is defined an OpenTelemetry endpoint, and Beyla was able to discover any process matching the entries in the discovery section. For best experience with generating service graph metrics, use a DNS for service discovery and make sure the DNS names match the OpenTelemetry service names used in Beyla. In Kubernetes environments, the OpenTelemetry service name set by the service name discovery is the best choice for service graph metrics.
• If the list contains application_process, the Beyla OpenTelemetry exporter exports metrics about the processes that run the instrumented application.
• If the list contains network, the Beyla OpenTelemetry exporter exports network-level metrics; but only if there is an OpenTelemetry endpoint defined. For network-level metrics options visit the network metrics configuration documentation.

This option affects the behaviour of the generation of application-level service graph metrics, which can be enabled by adding application_service_graph to the list of OpenTelemetry metric export features. By default, Beyla does not report application-level service graph metrics which are considered to be self-referencing. For example, self-references can be calls from local node metric scrape tools, or a service making an HTTP call to itself. Self-references not useful for the purpose of showing service graphs, while at the same time they increase the cardinality and the overall metric storage cost. To allow generation of application-level service graph metrics which also include self-references, change this option value to true.

A list of available instrumentations which are enabled, defined a comma separated list of strings. By default all available instrumentations are enabled, and you can choose to enable only some. The available instrumentations are as follows:

• * enables all instrumentations. If * is present in the list, the other values are simply ignored.
• http enables the collection of HTTP/HTTPS/HTTP2 application metrics.
• grpc enables the collection of gRPC application metrics.
• sql enables the collection of SQL database client call metrics.
• redis enables the collection of Redis client/server database metrics.
• kafka enables the collection of Kafka client/server message queue metrics.

For example, setting the instrumentations option to: http,grpc enables the collection of HTTP/HTTPS/HTTP2 and gRPC application metrics, while the rest of the instrumentations are be disabled.

The buckets object allows overriding the bucket boundaries of diverse histograms. See Overriding histogram buckets section for more details.

Specifies the default aggregation to use for histogram instruments.

Accepted values are:

• explicit_bucket_histogram (default): use Explicit Bucket Histogram Aggregation.
• base2_exponential_bucket_histogram: use Base2 Exponential Bucket Histogram Aggregation.

Overriding histogram buckets

For both OpenTelemetry and Prometheus metrics exporters, you can override the histogram bucket boundaries via a configuration file (see buckets YAML section of your metrics exporter configuration).

Sets the bucket boundaries for the metrics related to the request duration. Specifically:

• http.server.request.duration (OTEL) / http_server_request_duration_seconds (Prometheus)
• http.client.request.duration (OTEL) / http_client_request_duration_seconds (Prometheus)
• rpc.server.duration (OTEL) / rpc_server_duration_seconds (Prometheus)
• rpc.client.duration (OTEL) / rpc_client_duration_seconds (Prometheus)

If the value is unset, the default bucket boundaries follow the recommendation from the OpenTelemetry semantic conventions

Copy

0, 0.005, 0.01, 0.025, 0.05, 0.075, 0.1, 0.25, 0.5, 0.75, 1, 2.5, 5, 7.5, 10

Sets the bucket boundaries for the metrics related to request sizes. This is:

• http.server.request.body.size (OTEL) / http_server_request_body_size_bytes (Prometheus)
• http.client.request.body.size (OTEL) / http_client_request_body_size_bytes (Prometheus)

If the value is unset, the default bucket boundaries are:

Copy

0, 32, 64, 128, 256, 512, 1024, 2048, 4096, 8192

The default values are UNSTABLE and could change if Prometheus or OpenTelemetry semantic conventions recommend a different set of bucket boundaries.

Use native histograms and exponential histograms

For Prometheus, native histograms are enabled if you enable the native-histograms feature in your Prometheus collector.

For OpenTelemetry you can use exponential histograms for the predefined histograms instead of defining the buckets manually. You need to set up the standard OTEL_EXPORTER_OTLP_METRICS_DEFAULT_HISTOGRAM_AGGREGATION environment variable. See the histogram_aggregation section in the OTEL metrics exporter section for more information.

OTEL traces exporter

ℹ️ If you plan to use Beyla to send metrics to Grafana Cloud, consult the Grafana Cloud OTEL exporter for metrics and traces section for easier configuration.

YAML section otel_traces_export.

This component exports OpenTelemetry traces to a given endpoint. It will be enabled if its endpoint attribute is set (either via an YAML configuration file or via environment variables).

In addition to the properties exposed in this section, this component implicitly supports the environment variables from the standard OTEL exporter configuration.

Specifies the OpenTelemetry endpoint where the traces will be sent. If you plan to send the metrics directly to the Grafana Cloud OpenTelemetry endpoint, you might prefer to use the configuration options in the Using the Grafana Cloud OTEL endpoint to ingest metrics and traces section.

The OTEL_EXPORTER_OTLP_ENDPOINT environment variable sets a common endpoint for both the metrics and the traces exporters. The OTEL_EXPORTER_OTLP_TRACES_ENDPOINT environment variable or the endpoint YAML property, will set the endpoint only for the traces’ exporter node, so the metrics exporter won’t be activated unless explicitly specified.

According to the OpenTelemetry standard, if you set the endpoint via the OTEL_EXPORTER_OTLP_ENDPOINT environment variable, the OpenTelemetry exporter will automatically add the /v1/traces path to the URL. If you want to avoid this addition, you can use either the OTEL_EXPORTER_OTLP_TRACES_ENDPOINT environment variable or the environment YAML property to use exactly the provided URL without any addition.

A list of available instrumentations which are enabled, defined a comma separated list of strings. By default all available instrumentations are enabled, and you can choose to enable only some. The available instrumentations are as follows:

• * enables all instrumentations. If * is present in the list, the other values are simply ignored.
• http enables the collection of HTTP/HTTPS/HTTP2 application traces.
• grpc enables the collection of gRPC application traces.
• sql enables the collection of SQL database client call traces.
• redis enables the collection of Redis client/server database traces.
• kafka enables the collection of Kafka client/server message queue traces.

For example, setting the instrumentations option to: http,grpc enables the collection of HTTP/HTTPS/HTTP2 and gRPC application traces, while the rest of the instrumentations are be disabled.

Specifies the transport/encoding protocol of the OpenTelemetry traces endpoint.

The accepted values, as defined by the OTLP Exporter Configuration document are http/json, http/protobuf and grpc.

The OTEL_EXPORTER_OTLP_PROTOCOL environment variable sets a common protocol for both the metrics and the traces exporters. The OTEL_EXPORTER_OTLP_TRACES_PROTOCOL environment variable, or the protocol YAML property, will set the protocol only for the traces’ exporter node.

If this property is not provided, Beyla will guess it according to the following rules:

• Beyla will guess grpc if the port ends in 4317 (4317, 14317, 24317, …), as 4317 is the usual Port number for the OTEL GRPC collector.
• Beyla will guess http/protobuf if the port ends in 4318 (4318, 14318, 24318, …), as 4318 is the usual Port number for the OTEL HTTP collector.

Controls whether the OTEL client verifies the server’s certificate chain and host name. If set to true, the OTEL client accepts any certificate presented by the server and any host name in that certificate. In this mode, TLS is susceptible to a man-in-the-middle attacks. This option should be used only for testing and development purposes.

Sampling policy

Beyla accepts the standard OpenTelemetry environment variables to configure the sampling ratio of traces.

In addition, you can configure the sampling under the sampler YAML subsection of the otel_traces_export section. For example:

yaml Copy

otel_traces_export:
  sampler:
    name: "traceidratio"
    arg: "0.1"

If you are using the Grafana Alloy as your OTEL collector, you can configure the sampling policy at that level instead.

Specifies the name of the sampler. It accepts the following standard sampler names from the OpenTelemetry specification:

• always_on: samples every trace. Be careful about using this sampler in an application with significant traffic: a new trace will be started and exported for every request.
• always_off: samples no traces.
• traceidratio: samples a given fraction of traces (specified by the arg property that is explained below). The fraction must be a real value between 0 and 1. For example, a value of "0.5" would sample 50% of the traces. Fractions >= 1 will always sample. Fractions < 0 are treated as zero. To respect the parent trace’s sampling configuration, the parentbased_traceidratio sampler should be used.
• parentbased_always_on (default): parent-based version of always_on sampler (see explanation below).
• parentbased_always_off: parent-based version of always_off sampler (see explanation below).
• parentbased_traceidratio: parent-based version of traceidratio sampler (see explanation below).

Parent-based samplers are composite samplers which behave differently based on the parent of the traced span. If the span has no parent, the root sampler is used to make sampling decision. If the span has a parent, the sampling configuration would depend on the sampling parent.

Specifies the argument of the selected sampler. Currently, only traceidratio and parentbased_traceidratio require an argument.

In YAML, this value MUST be provided as a string, so even if the value is numeric, make sure that it is enclosed between quotes in the YAML file, (for example, arg: "0.25").

Filter metrics and traces by attribute values

You might want to restrict the reported metrics and traces to very concrete event types based on the values of the attributes (for example, filter network metrics to report only TCP traffic).

The filter YAML section allows filtering both application and network metrics by attribute values. It has the following structure:

yaml Copy

filter:
  application:
    # map of attribute matches to restrict application metrics
  network:
    # map of attribute matches to restrict network metrics

For a list of metrics under the application and network family, as well as their attributes, check the Beyla exported metrics document.

Each application and network filter section is a map where each key is an attribute name (either in Prometheus or OpenTelemetry format), with either the match or the not_match property. Both properties accept a glob-like string (it can be a full value or include wildcards). If the match property is set, Beyla only reports the metrics and traces matching the provided value for that given attribute. The not_match property is the negation of match.

The following example reports network metrics for connections targeting the destination port 53, excluding the UDP protocol:

yaml Copy

filter:
  network:
    transport:
      not_match: UDP
    dst_port:
      match: "53"

Using the Grafana Cloud OTEL endpoint to ingest metrics and traces

You can use the standard OpenTelemetry variables to submit the metrics and traces to any standard OpenTelemetry endpoint, including Grafana Cloud.

Alternatively, Beyla can be configured to submit OpenTelemetry data to the Grafana Cloud OTEL endpoint using its own custom variables, allowing an easier setup of the endpoint and the authentication.

The properties can be defined via environment variables, or under the grafana top-level YAML section, otlp subsection. For example:

yaml Copy

grafana:
  otlp:
    cloud_zone: prod-eu-west-0
    cloud_instance_id: 123456

Accepts a list of strings with the kind of data that will be submitted to the OTLP endpoint. It accepts metrics and/or traces as values.

The cloud zone of your Grafana endpoint. This will be used to compose the Grafana OTLP URL. For example, if the value is prod-eu-west-0, the used OTLP URL will be https://otlp-gateway-prod-eu-west-0.grafana.net/otlp.

If any of the OTEL_EXPORTER_OTLP_ENDPOINT, OTEL_EXPORTER_OTLP_METRICS_ENDPOINT or OTEL_EXPORTER_OTLP_TRACES_ENDPOINT variables are defined, they will override the destination endpoint, so the cloud_zone configuration option will be ignored.

Your Grafana user name. It is usually a number but it must be set as a string inside the YAML file.

API key of your Grafana Cloud account.

Prometheus HTTP endpoint

ℹ️ The Prometheus scraper might override the values of the instance and job labels. To keep the original values as set by Beyla, make sure to configure the Prometheus scraper to set the honor_labels option to true.

(How to set honor_labels in Grafana Alloy).

YAML section prometheus_export.

This component opens an HTTP endpoint in the auto-instrumentation tool that allows any external scraper to pull metrics in Prometheus format. It is enabled if the port property is set.

Specifies the HTTP port for the Prometheus scrape endpoint. If unset or 0, no Prometheus endpoint is open.

Specifies the HTTP query path to fetch the list of Prometheus metrics.

The group of attributes for a metric instance is not reported anymore if the time since the last update is greater than this Time-To-Leave (TTL) value.

The purpose of this value is to avoid reporting indefinitely finished application instances.

The buckets object allows overriding the bucket boundaries of diverse histograms. See Overriding histogram buckets section for more details.

A list of metric groups that are allowed to be exported. Each group belongs to a different feature of Beyla: application-level metrics or network metrics.

• If the list contains application, the Beyla Prometheus exporter exports application-level metrics; but only if the Prometheus port property is defined, and Beyla was able to discover any process matching the entries in the discovery section.
• If the list contains application_span, the Beyla Prometheus exporter exports application-level metrics in traces span metrics format; but only if the Prometheus port property is defined, and Beyla was able to discover any process matching the entries in the discovery section.
• If the list contains application_service_graph, the Beyla Prometheus exporter exports application-level service graph metrics; but only if the Prometheus port property is defined, and Beyla was able to discover any process matching the entries in the discovery section. For best experience with generating service graph metrics, use a DNS for service discovery and make sure the DNS names match the OpenTelemetry service names used in Beyla. In Kubernetes environments, the OpenTelemetry service name set by the service name discovery is the best choice for service graph metrics.
• If the list contains application_process, the Beyla Prometheus exporter exports metrics about the processes that run the instrumented application.
• If the list contains network, the Beyla Prometheus exporter exports network-level metrics; but only if the Prometheus port property is defined. For network-level metrics options visit the network metrics configuration documentation.

This option affects the behaviour of the generation of application-level service graph metrics, which can be enabled by adding application_service_graph to the list of Prometheus metric export features. By default, Beyla does not report application-level service graph metrics which are considered to be self-referencing. For example, self-references can be calls from local node metric scrape tools, or a service making an HTTP call to itself. Self-references not useful for the purpose of showing service graphs, while at the same time they increase the cardinality and the overall metric storage cost. To allow generation of application-level service graph metrics which also include self-references, change this option value to true.

A list of available instrumentations which are enabled, defined a comma separated list of strings. By default all available instrumentations are enabled, and you can choose to enable only some. The available instrumentations are as follows:

• * enables all instrumentations. If * is present in the list, the other values are simply ignored.
• http enables the collection of HTTP/HTTPS/HTTP2 application metrics.
• grpc enables the collection of gRPC application metrics.
• sql enables the collection of SQL database client call metrics.
• redis enables the collection of Redis client/server database metrics.
• kafka enables the collection of Kafka client/server message queue metrics.

For example, setting the instrumentations option to: http,grpc enables the collection of HTTP/HTTPS/HTTP2 and gRPC application metrics, while the rest of the instrumentations are be disabled.

Internal metrics reporter

YAML section internal_metrics.

This component reports certain internal metrics about the behavior of the auto-instrumentation tool. Currently, only Prometheus export is supported. It is enabled if the internal_metrics section contains a prometheus subsection with the port property set.

Example:

yaml Copy

internal_metrics:
  prometheus:
    port: 6060
    path: /internal/metrics

Specifies the HTTP port for the Prometheus scrape endpoint. If unset or 0, no Prometheus endpoint is open and no metrics are accounted.

Its value can be the same as prometheus_export.port (both metric families share the same HTTP server, though they can be accessed in different paths), or a different value (two different HTTP servers for the different metric families).

Specifies the HTTP query path to fetch the list of Prometheus metrics. If prometheus_export.port and internal_metrics.prometheus.port have the same values, this internal_metrics.prometheus.path value can be different from prometheus_export.path, to keep both metric families separated, or the same (both metric families are listed in the same scrape endpoint).

YAML file example

yaml Copy

open_port: 443
service_name: my-instrumented-service
log_level: DEBUG

ebpf:
  wakeup_len: 100

otel_traces_export:
  endpoint: https://otlp-gateway-prod-eu-west-0.grafana.net/otlp

prometheus_export:
  port: 8999
  path: /metrics