Telemetry Agents

Telemetry agents sit close to the systems that produce data. They read container logs, application logs, host metrics, Prometheus metrics, or OTLP signals, then forward that data to Parseable with the labels and metadata you need for search and dashboards.

Parseable supports ingesting logs, metrics, and traces from agents that teams already use in production. Plain JSON logs are sent to /api/v1/ingest with an X-P-Stream header. OTLP logs, metrics, and traces are sent to /v1/logs, /v1/metrics, and /v1/traces; Parseable derives the signal type from the endpoint and uses X-P-Stream for the destination dataset. For OTLP examples, send the Parseable API key as a Bearer token in the Authorization header.

Use this section when you already have an agent running, or when you want to pick one for a new deployment.

Agent	Good fit
Apache Log4j 2	Java application logs sent directly from the app over HTTP.
Filebeat	Existing Beats pipelines where Filebeat ships logs to Logstash, and Logstash forwards them to Parseable.
Fluent Bit	Lightweight container, Kubernetes, host, and edge log collection.
Fluentd	Existing Fluentd pipelines that need an HTTP destination for Parseable.
Logstash	Filter-heavy log pipelines and Beats ingestion before sending data to Parseable.
OpenTelemetry Collector	Logs, metrics, and traces pipelines built around OpenTelemetry receivers, processors, and exporters.
Prometheus	Metrics ingestion through Prometheus Remote Write.
Promtail	Existing Loki-style log pipelines that can be routed through the OpenTelemetry Collector.
Syslog-ng	Network, host, and appliance logs collected through syslog.
Vector	High-throughput log pipelines with transforms before HTTP delivery.

If you are starting fresh on Kubernetes, Fluent Bit or the OpenTelemetry Collector is usually the simplest place to begin. If your team already runs Filebeat, Fluentd, Logstash, Promtail, Syslog-ng, or Vector, keep that pipeline and update only the output stage that sends data to Parseable.