Does Dora-Dataflow not support multiple Dynamic inputs?

[chengzi0103]

Describe the bug When I use multiple Dynamic nodes in a Dataflow, the Dataflow cannot capture the Dynamic inputs after starting (no response). Also, when I run python dynamic.py, there is no response.

I would like to ask, is this a bug or the current design, or is the Dynamic structure designed this way? Will this issue be fixed in future versions?

[Hennzau]

Describe the bug

When I use multiple Dynamic nodes in a Dataflow, the Dataflow cannot capture the Dynamic inputs after starting (no response). Also, when I run python dynamic.py, there is no response.

I would like to ask, is this a bug or the current design, or is the Dynamic structure designed this way? Will this issue be fixed in future versions?

Hi! Thanks for the report,

• what is exactly the version of Dora you are using ？ crates.io 0.3.5, main branch ? Fix tag ?

• can you give us the exact dataflow description (YAML) ?

• can you give us the python code of your dynamic nodes ?

[chengzi0103]

Describe the bug When I use multiple Dynamic nodes in a Dataflow, the Dataflow cannot capture the Dynamic inputs after starting (no response). Also, when I run python dynamic.py, there is no response. I would like to ask, is this a bug or the current design, or is the Dynamic structure designed this way? Will this issue be fixed in future versions?

Hi! Thanks for the report,

• what is exactly the version of Dora you are using ？ crates.io 0.3.5, main branch ? Fix tag ?
• can you give us the exact dataflow description (YAML) ?
• can you give us the python code of your dynamic nodes ?

@Hennzau

1. We are using version 3.5 of dora-rs, and the version is the main branch.
2. We used the dynamic method for nodes with node_id names web_search_task and reasoner_task_input.
3. task_input.py is our Python code.

nodes:
  - id: web_search_task
    path: dynamic
    inputs:
      direction: dora/timer/secs/1
    outputs:
    - task
  - id: web_search_agent
    operator:
      python: ./scripts/web_search_agent.py
      inputs:
        web_search_task: web_search_task/task
      outputs:
      - web_search_results
      - web_search_resource
      - web_search_aggregate_output
  - id: more_question_agent
    operator:
      python: ./scripts/more_question_agent.py
      inputs:
        web_search_aggregate_output: web_search_agent/web_search_aggregate_output
      outputs:
      - more_question_results
      - web_search_aggregate_output
  - id: web_search_output
    operator:
      python: ./scripts/web_search_output.py
      inputs:
        web_search_aggregate_output: more_question_agent/web_search_aggregate_output
      outputs:
      - web_search_output
      - web_search_results
      - web_search_resource
      - more_question_results
      - web_search_task
  - id: reasoner_task_input
    path: dynamic
    inputs:
      direction: dora/timer/secs/1
    outputs:
    - reasoner_task
  - id: reasoner_agent
    operator:
      python: ./scripts/reasoner_agent.py
      inputs:
        reasoner_task: reasoner_task_input/reasoner_task
        more_question_results: more_question_agent/more_question_results
      outputs:
      - reasoner_result
  - id: reasoner_output
    operator:
      python: ./scripts/reasoner_output.py
      inputs:
        reasoner_result: reasoner_agent/reasoner_result
      outputs:
      - reasoner_output

Apart from different node_ids, other results of the dynamic nodes are the same.

import json

from dora import Node
import pyarrow as pa

node = Node("web_search_task")

event = node.next()
task_data = input('Please enter your task:  ',)
node.send_output('task',pa.array([json.dumps(task_data)]),event["metadata"])

[Hennzau]

Hi, sorry for the long time between my last response. Did you succeed in receiving multiple inputs?

[chengzi0103]

Hi, sorry for the long time between my last response. Did you succeed in receiving multiple inputs?

Thank you so much, we have successfully resolved the issue