rhythmic-hamburger-56597
11/20/2024, 11:00 AMdomain = "development"
staging_project = "kaas-pipeline-4375a9fa-74e3-4897-b4ab-5655fef1c8c3"
task_name = 'utils.database.map_write_table_chunks_0d5e41499a01d186dfae95c68432296e-arraynode'
version = 'v3.0.6'
flyte_task = staging_remote.fetch_task(name=task_name, version=version, project='kaas-pipeline')
# Recursive parser to interpret Flyte types
def parse_flyte_type(literal_type: LiteralType) -> str:
if literal_type.collection_type:
return f"List[{parse_flyte_type(literal_type.collection_type)}]"
elif literal_type.map_value_type:
return f"Dict[str, {parse_flyte_type(literal_type.map_value_type)}]"
elif literal_type.blob:
blob_format = literal_type.blob.format or "UnknownFormat"
return f"Blob(format={blob_format})"
elif literal_type.simple:
return literal_type.simple
return "Unknown"
# Get type hints for task inputs
def get_task_type_hints(task):
type_hints = {}
if task and task.interface and task.interface.inputs:
for input_name, input_variable in task.interface.inputs.items():
type_hints[input_name] = parse_flyte_type(input_variable.type)
return type_hints
# Fetch and display type hints
type_hints = get_task_type_hints(flyte_task)
fetched_input = staging_remote.get("<flyte://v1/kaas-pipeline-4375a9fa-74e3-4897-b4ab-5655fef1c8c3/development/lp-61494237-7542-439e-b170-357bed12b14d/n2-0-n4/i>")
execution = staging_remote.execute(entity=flyte_task, inputs=fetched_input.literals, type_hints = type_hints)
got the below error at staging_remote.execute(...)
FlyteSystemException: SYSTEM:Unknown: error=None, cause=<_InactiveRpcError of RPC that terminated with:
status = StatusCode.INTERNAL
details = "failed to create workflow in propeller Request entity too large: limit is 3145728"
debug_error_string = "UNKNOWN:Error received from peer ipv4:3.130.23.25:443 {created_time:"2024-11-20T16:20:14.689298+05:30", grpc_status:13, grpc_message:"failed to create workflow in
propeller Request entity too large: limit is 3145728"}"
help me resolve this errorjolly-florist-66728
11/20/2024, 11:00 AMmax-output-size-bytes
parameter in the flyte-propeller-config
ConfigMap. This requires access to the Flyte deployment configuration.
2. Reduce Input Size: If possible, try to reduce the size of the input data being sent to the task. This might involve breaking down the data into smaller chunks or optimizing the data format.
3. Use External Storage: For large data inputs, consider using external storage solutions like S3 or GCS. You can upload the data to a storage bucket and pass the reference (e.g., a URL) to the task instead of the data itself.
Would you like more detailed guidance on any of these approaches?
Sources:
- Flyte Community Discussion on Output Size 👍 0 button 👎 0 button 💡 Hint: Mention @runllm in the thread for followups.average-finland-92144
11/22/2024, 8:43 PMrhythmic-hamburger-56597
11/23/2024, 8:57 AM