Security and Authentication

Cluster Metadata Collection

We collect non-sensitive data on the cluster that helps us understand how Runhouse is being used. This data includes:

• Python version

• Resources (cpus, gpus, memory)

• Cloud provider

• Region

• Instance type

Performance & Error Monitoring

We collect data from normal package usage that helps us understand the performance and errors that are raised. This data is collected and sent to Sentry, a third-party error tracking service.

Removing Collected Data

If you would like us to remove your collected data, please contact the Runhouse team (first name at run.house)

Disabling Data Collection

To disable data collection and error tracking collection, set the environment variable DISABLE_DATA_COLLECTION to True. Alternatively, set disable_data_collection to true in your local Runhouse config (~/.rh/config.yaml), or in Python:

import runhouse as rh
rh.configs.disable_data_collection()

Cluster Observability

Runhouse collects various telemetry data by default on clusters. This data will be used to provide better observability into logs, traces, and metrics associated with clusters. We will not sell data or buy any observability data collected.

To disable observability globally for all clusters, set the environment variable disable_observability to True. Alternatively, set disable_observability to true in your local Runhouse config (~/.rh/config.yaml), or in Python:

import runhouse as rh
rh.configs.disable_observability()