lhm-engine-bind — Validation Engine Query and Binding

Positioning

add-data-check-task (POP/CLI) has no engine parameter; for tasks created via the API, srcEngineId/dstEngineId are always empty. The backend then falls back to "using the data source itself as the engine"; in MaxCompute scenarios a [RuntimeException]No default project specified. is thrown at the worker stage, and it is only visible in the job/step's err_message.

update-data-check-task supports engine parameters; this skill uses it to fill the gap.

Steps

1. Query Available Engines

python scripts/run.py list --ds-type MaxCompute
python scripts/run.py list --ds-type MaxCompute --check-type 1   # filter by the engine types allowed by the template

2. Bind Engines to an Existing Task

python scripts/run.py bind --task-id 10001 \
  --src-engine-id 456 --src-engine-name mc_engine --src-engine-type MaxCompute \
  --dst-engine-id 456 --dst-engine-name mc_engine --dst-engine-type MaxCompute

After binding, exec-data-check-save-task must be rerun to generate a new batch before executing: both the batch and the table configuration copy the engine fields from the task record as it stood at that time.

3. Read Back to Confirm

python scripts/run.py show --task-id 10001

The binding is successful only when src_engine_id / dst_engine_id are non-empty.

Pitfalls

• Ordering: for new tasks, the engine must be bound before add-data-check-config (the run_* functions already do this internally); when fixing an existing task, save again after binding.
• list uses the 2022-11-15 version API (list-meta-data-component-engine); aliyun_cli has the version preset built in, no manual specification needed.
• When the engine id equals the data source id, binding does not change execution behavior; if No default project specified. is still reported, the data source itself lacks a default project — it must be completed in the console; binding an engine is ineffective.
• Outdated plugin: if update-data-check-task does not recognize the engine flags, aliyun_cli explicitly reports CLI_PARAM_UNSUPPORTED instead of silently dropping the parameters; upgrade the plugin in that case.

Verification

python scripts/run.py list --ds-type MaxCompute     # candidates can be listed
python scripts/run.py show --task-id <id>           # engine fields are non-empty