When running edr report as part of a Databricks workflow, if the command hangs indefinitely (neither succeeding nor failing), there are a few common causes and troubleshooting steps based on similar issues reported by the community: 1. Permissions and Working Directory: Elementary tries to create a target directory (like edr_target) for its internal files. If the current working directory is not writable (e.g., inside a cloned repo or a restricted path in Databricks), this can cause the process to hang or fail. Solution: Before running any edr command, change the working directory to a location that is writable (such as /tmp). This ensures Elementary can create its required folders and files. Example workaround:

python    import os    os.chdir("/tmp")  # Switch to a writable directory before running edr    # Now run your edr command

This approach was confirmed to resolve similar issues in Databricks Workflows, as the Config class in Elementary tries to create directories before command-line arguments like --target-path are processed. If you don't switch to a writable directory, the process may hang or error out early in execution [Databricks Workflows directory issue](https://slack.com/archives/C02CTC89LAX/p1740037553539379). 2. Cluster Access Mode and Permissions: If you are running on a Databricks cluster in "shared" access mode, you may encounter permission errors when Elementary tries to write to certain locations (like the Python site-packages directory). This can cause the process to hang or fail. Solution: Use a "single user" access mode cluster, or ensure that the paths Elementary needs to write to are accessible. See: [Databricks shared cluster permission error](https://github.com/elementary-data/elementary/issues/1584). 3. Resource Constraints or Large Data Volumes: If the report generation is processing a large amount of data (many test results, long days-back window), it may appear to hang due to long-running queries or memory issues. Solution: Try reducing the --days-back or --executions-limit parameters to limit the data processed. This can help determine if the issue is due to data volume [Timeouts and large data](https://slack.com/archives/C02CTC89LAX/p1691169511715969). 4. Check for Silent Errors: Sometimes, errors are not surfaced in the main output but are present in the edr.log file. Solution: Check the edr.log file in your target directory for any hidden errors or stack traces that might indicate the root cause [Checking edr.log](https://slack.com/archives/C02CTC89LAX/p1697659818900659). Summary of steps to try: - Change the working directory to a writable location (e.g., /tmp) before running edr. - Ensure your cluster access mode and permissions allow writing to the necessary paths. - Reduce the data processed by limiting --days-back or --executions-limit. - Check edr.log for additional error details. If after these steps the issue persists, please provide the edr.log output for further troubleshooting.