Ingest telemetry data
Before you can use Chronosphere Observability Platform to view and manage your observability data (of whatever type), you need to get that data into Observability Platform.
Observability Platform supports multiple methods to ingest telemetry data, which depend on which type of data you want to ingest:
| Ingestion method | Change events | Logs | Metrics | Traces |
|---|---|---|---|---|
| Chronosphere Collector | ||||
| OpenTelemetry Collector | ||||
| Telemetry Pipeline or Core Agent | ||||
| Existing pipeline | ||||
| Direct to Observability Platform |
After ingesting telemetry data, you can use the control mechanisms that Observability Platform provides to control costs and ensure you're ingesting only the data you care about.
Chronosphere utilizes push and pull models of ingestion, depending on the data collected and the method of ingestion.
- Pull models, like the Chronosphere Collector, scrape data from external sources and pull it in. These metrics have consistent reporting intervals.
- Push models, like tracing, send data to Chronosphere, which is then processed. These metrics can have a broad spectrum of reporting frequency, from large bursts of data to long periods with no data reporting.
| Telemetry | Push | Pull |
|---|---|---|
| Metrics | ||
| Traces | ||
| Logs |
The ingestion model depends on the data source. Metrics pushed to Observability Platform can have latency delays or sparse time series which can result in unexpected query results.
Metric and trace data
To ingest metrics and traces, you must configure one of the collection methods. Use the Chronosphere Collector or OpenTelemetry Collector for metric and trace data. Google Cloud metrics can also be directly ingested.
Change event data
Change event data gets sent directly to Observability Platform, and doesn't require an independent collection mechanism. Use Observability Platform or CreateEvent API to send your change events (opens in a new tab).
You can also ingest change events directly from supported third-party tools such as Amazon EventBridge, Buildkite, and LaunchDarkly.
Log data
To ingest log data, Chronosphere recommends using Core Agent or a combination of Core Agent plus Telemetry Pipeline. Chronosphere also supports using your existing ingest pipeline to ingest log data, such as Fluent Bit or LogStash. You can also ingest log data using the OpenTelemetry Collector.
Control costs
Ingesting more telemetry data increases your overall data and storage costs. Observability Platform provides several control mechanisms to manage the data you ingest.
Metric data
Use the following methods to control the metric data you ingest:
- Profile and analyze metrics for their usage and cost.
- Create drop rules to drop incoming metrics based on labels.
- Shape metric data with quotas.
- Reduce the persisted metrics and cardinality at the cost of increased writes.
Tracing data
Use the following methods to control the tracing data you ingest:
- Create sampling rules to control costs and maximize the usefulness of your tracing data.
- Create datasets to map sets of traces to named groups relevant to your organization, and then track processed and persisted bytes for those groups over time.
- Apply behaviors to your datasets to set sampling rates for your datasets without needing to write fine-grained sampling rules.