sdods run -p billing -e staging -l ui --artifacts-dir /var/tmp/sdods-runssdods heal report --lastNo run found.
Run `sdods run` first or pass --run <id>.Exit code 2. The run directory is there and full of files — I can open heal.jsonl inside it.
sdods run -p billing -e staging -l ui --artifacts-dir /var/tmp/sdods-runssdods heal report --lastNo run found.
Run `sdods run` first or pass --run <id>.Exit code 2. The run directory is there and full of files — I can open heal.jsonl inside it.
Accepted answer
heal report resolves the artifacts directory from SDODS_ARTIFACTS_DIR or the default .sdods/runs under the repo root. It knows nothing about the --artifacts-dir you passed to run, so it looked in the default location, found nothing and raised RUN_FAILED with exit code 2.
Either point both commands at the same place:
SDODS_ARTIFACTS_DIR=/var/tmp/sdods-runs sdods heal report --lastor leave the runs where they land and name the one you want:
sdods heal report --run <id>There is a second cause that produces the same message even when the path is right. --last picks the most recent subdirectory that contains a run.json. A directory of scenario output with no manifest at the top is invisible to it, which is what you get when something kills the run before it writes one.
Have a different question? Ask it.