[BUG] Test-connect to demo DB failed
likawind opened this issue · comments
Wei Chen commented
Describe the bug
The UI showed failure when test-connecting to demo DB.
stack trace from notebook:
AqueductError Traceback (most recent call last)
/tmp/ipykernel_83/3226702849.py in <module>
2 raw_data = demo_db.sql("select * from house_prices;")
3
----> 4 filled_data = fill_missing_data(raw_data)
5 encoded_data = encode_labels(filled_data)
6 unskewed_data = unskew_features(encoded_data)
/opt/conda/lib/python3.9/site-packages/aqueduct/decorator.py in wrapped(*input_artifacts)
474
475 def wrapped(*input_artifacts: BaseArtifact) -> Union[BaseArtifact, List[BaseArtifact]]:
--> 476 return _wrapped_util(*input_artifacts, execution_mode=ExecutionMode.EAGER)
477
478 # Enable the .local(*args) attribute, which calls the original function with the raw inputs.
/opt/conda/lib/python3.9/site-packages/aqueduct/decorator.py in _wrapped_util(execution_mode, *input_artifacts)
461 file=zip_file,
462 )
--> 463 return wrap_spec(
464 OperatorSpec(
465 function=function_spec,
/opt/conda/lib/python3.9/site-packages/aqueduct/decorator.py in wrap_spec(spec, op_name, output_artifact_type_hints, description, execution_mode, *input_artifacts)
151 if execution_mode == ExecutionMode.EAGER:
152 # Issue preview request since this is an eager execution.
--> 153 output_artifacts = artifact_utils.preview_artifacts(dag, output_artifact_ids)
154 else:
155 # We are in lazy mode.
/opt/conda/lib/python3.9/site-packages/aqueduct/artifacts/utils.py in preview_artifacts(dag, target_artifact_ids, parameters)
65 )
66
---> 67 preview_resp = globals.__GLOBAL_API_CLIENT__.preview(dag=subgraph)
68
69 # Process all the target artifacts first. Assumption: the preview response contains a result entry
/opt/conda/lib/python3.9/site-packages/aqueduct/api_client.py in preview(self, dag)
378
379 preview_resp = self._construct_preview_response(resp)
--> 380 _handle_preview_resp(preview_resp, dag)
381 return preview_resp
382
/opt/conda/lib/python3.9/site-packages/aqueduct/api_client.py in _handle_preview_resp(preview_resp, dag)
129
130 failure_err_msg = "\n".join(op_err_msgs)
--> 131 raise AqueductError(f"Preview Execution Failed:\n\n{failure_err_msg}\n")
132
133
AqueductError: Preview Execution Failed:
Operator `fill_missing_data` failed!
Unable to get object from storage
Object does not exist in storage.
ORIGINAL STACK TRACE:
github.com/aqueducthq/aqueduct/lib/storage.init
github.com/aqueducthq/aqueduct/lib/storage/storage.go:11 +0xfb82dd
runtime.doInit
runtime/proc.go:6265 +0x45062b
runtime.doInit
runtime/proc.go:6242 +0x4505b1
runtime.doInit
runtime/proc.go:6242 +0x4505b1
runtime.doInit
runtime/proc.go:6242 +0x4505b1
runtime.doInit
runtime/proc.go:6242 +0x4505b1
runtime.doInit
runtime/proc.go:6242 +0x4505b1
runtime.doInit
runtime/proc.go:6242 +0x4505b1
runtime.doInit
runtime/proc.go:6242 +0x4505b1
runtime.main
runtime/proc.go:208 +0x442a24
runtime.goexit
runtime/asm_amd64.s:1371 +0x479960
Sorry, we've run into an unexpected error! Please create bug report in github: https://github.com/aqueducthq/aqueduct/issues/new?assignees=&labels=bug&template=bug_report.md&title=%5BBUG%5D . We will get back to you as soon as we can.
To Reproduce
Notebook script:
demo_db = client.integration("aqueduct_demo")
raw_data = demo_db.sql("select * from house_prices;")
filled_data = fill_missing_data(raw_data)
encoded_data = encode_labels(filled_data)
unskewed_data = unskew_features(encoded_data)
featurized_data = one_hot_encode(unskewed_data)
predictions = predict(raw_data, featurized_data)
df = predictions.get()
df[["PredictedSalePrice"]]
Expected behavior
N/A
Desktop (please complete the following information):
N/A pending more updates from original bug reporter
Additional context
Need to know engine / cluster settings
Wei Chen commented
We were able to figure the root-cause being the the server not having the permission to access certain files. We take it as resolved as no further action / fix is required.