Last updated:
Datasets
Datasets provide a structure to separate unrelated observability data while keeping organizational data like users, notification settings, and billing information.
What are datasets?
Datasets are used in Dash0 to separate observability information within one organization. Data that should be correlated must live in one dataset, while totally unrelated data can live in different datasets (e.g., prod—and pre-prod environments).
The most common use cases for datasets are…
- Environments, i.e., datasets called
productionandtest. - Auditing logging.
- Local development and experiments.
Managing datasets
Users can create multiple datasets in the "manage" area of your Dash0 environment.
By default, Dash0 comes with two datasets:
- Default: This is your dataset that is automatically created and that contains your observability information.
- OTelDemo: Contains data from the OpenTelemetry Demo environment. It enables you to try Dash0 when you have not sent data to Dash0 yet.
Identifier
Every dataset has an Identifier. Telemetry and queries use it to name the dataset.
When you create a dataset, Dash0 derives the identifier from the dataset name. You can change it before you save. After that, you cannot change it.
Deleting and restoring datasets
You can delete a dataset in Settings → Datasets. The default dataset cannot be deleted.
Dash0 keeps the assets of a deleted dataset for 30 days. This includes dashboards, views, check rules, synthetic checks, recording rules, SLOs, and time series aggregations. Spam filters are not kept. After 30 days, Dash0 removes the dataset, its permissions, and those assets. Telemetry is removed as its retention expires.
To restore a deleted dataset, create a dataset with the same identifier within the 30 days. The dataset comes back with its assets. Its settings, such as retention, come from the new dataset.
Datasets managed with Terraform
You cannot edit the settings of a dataset created with the Terraform provider in the UI. This keeps the dataset from drifting from your configuration. Change its settings in your Terraform configuration.
Sending Data to a specific Dataset
When sending telemetry to Dash0, use the Dash0-Dataset header to select the dataset. This is supported for almost all Dash0 ingestion endpoints. You can find usage instructions within Dash0's onboarding screens and the integrations hub.
The auth token must be permitted to write to the dataset you name, otherwise Dash0 rejects the request. When you omit the header, Dash0 writes the telemetry to the token's default ingestion dataset.
Further Reading
- About Access Control — Flexible access control system for team collaboration
- About Organizations — Top-level entity grouping users, datasets, and billing
- Endpoints — Public-facing endpoints for reading, configuring, and ingesting data
- Redact Cardholder Data — Opt-in dataset setting that masks payment card numbers before storage
- Semantic Convention Upgrades — Automatic schema migrations for consistent telemetry data