lhm-poll-status

Polls the execution status of an LHM data validation task until the task reaches a terminal state (STOPPED / FAILED / FINISHED) or the timeout is reached.

Steps

1. Parse the command-line arguments (task-id, batch-id, region, timeout-sec, interval-sec, etc.).
2. Call build_client_from_config() to build the LHM client (the network layer goes through the aliyun CLI, read from data_validation_config.yaml, with CLI arguments taking precedence).
3. Call poll_exec_status() to poll the task status; redirect sys.stdout to capture the log prints inside the common functions and avoid polluting stdout.
4. If the status reaches a terminal state, output {"ok": true, "exec_status": <int>, "exec_status_text": "..."}.
5. If it times out, output {"ok": false, "exec_status": -1, "reason": "timeout"} and exit(1).

Pitfalls

• --task-id and --batch-id must be provided together; missing either one causes the API query to fail.
• Long-running batch tasks may require increasing --timeout-sec; the default of 1800 seconds may not be enough.
• Credential precedence: command-line arguments > data_validation_config.yaml (~/.lhm/data_validation_config.yaml) > environment variables.
• The common functions use print internally for polling logs; the script captures them via io.StringIO and writes them to stderr, so stdout contains only the final JSON.

Verification

• For a normally completed batch, stdout should output something like:

  {"ok": true, "exec_status": 4, "exec_status_text": "FINISHED"}
  

• For the timeout scenario, stderr should output:

  {"ok": false, "exec_status": -1, "reason": "timeout"}
  

• Running python -m py_compile atomic-skills/lhm-poll-status/scripts/run.py produces no syntax errors.

Input/Output Examples

python atomic-skills/lhm-poll-status/scripts/run.py \
  --task-id 12345 \
  --batch-id 67890 \
  --region hangzhou \
  --timeout-sec 1800 \
  --interval-sec 5

Successful output:

{"ok": true, "exec_status": 4, "exec_status_text": "FINISHED"}