We use cookies

We use cookies to ensure you get the best experience on our website. For more information on how we use cookies, please see our cookie policy.

By clicking "Accept", you agree to our use of cookies.
Learn more.

Self HostingTrace Sampling

Trace Sampling

For a very high-volume setup, it may be desirable to sample results for the dashboard for the purpose of limiting the amount of data stored in the Hatchet database. This does not impact the behavior of the Hatchet engine and all tasks will still be processed. This can be done by setting the following environment variables:

SERVER_SAMPLING_ENABLED=t
SERVER_SAMPLING_RATE=0.1 # only 10% of results will be sampled

Sampling is done at the workflow run level, so all tasks within the same workflow will be sampled, along with all of their events. Sampling has the following limitations:

  • Parent tasks which spawn child tasks are not guaranteed to be sampled, even if their children are. This means that the child task may be shown in the dashboard without a corresponding parent task, and vice versa.
  • There is no way to configure sampling to ensure that failure events are sampled.
  • Only tasks which are sampled can be cancelled or replayed via the REST APIs: do not use this feature if dependent on programmatic cancellations and replays.
Read Replicas