HTTPStatusError("Client error '422 Unprocessable Entity' ...

[wkdghdus]

Hi, there are some issues trying to implement this template to the project I am working on.

When I import this project straight into my project without any modification, it returns such error.

HTTPStatusError("Client error '422 Unprocessable Entity' for url 'http://api/threads/c7ff1dda-9fb2-4aac-aaf2-c77598d66938/runs'\nFor more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/422")Traceback (most recent call last):


  File "/usr/local/lib/python3.11/site-packages/langgraph/pregel/__init__.py", line 1511, in astream
    async for _ in runner.atick(


  File "/usr/local/lib/python3.11/site-packages/langgraph/pregel/runner.py", line 194, in atick
    _panic_or_proceed(


  File "/usr/local/lib/python3.11/site-packages/langgraph/pregel/runner.py", line 273, in _panic_or_proceed
    raise exc


  File "/usr/local/lib/python3.11/site-packages/langgraph/pregel/retry.py", line 102, in arun_with_retry
    await task.proc.ainvoke(task.input, config)


  File "/usr/local/lib/python3.11/site-packages/langgraph/utils/runnable.py", line 453, in ainvoke
    input = await asyncio.create_task(coro, context=context)
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^


  File "/usr/local/lib/python3.11/site-packages/langgraph/utils/runnable.py", line 236, in ainvoke
    ret = await asyncio.create_task(coro, context=context)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^


  File "/deps/__outer_interiorGPT/src/squads/conversationTeam/chatbot/graph.py", line 50, in schedule_memories
    await memory_client.runs.create(


  File "/usr/local/lib/python3.11/site-packages/langgraph_sdk/client.py", line 1474, in create
    return await self.http.post(f"/threads/{thread_id}/runs", json=payload)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^


  File "/usr/local/lib/python3.11/site-packages/langgraph_sdk/client.py", line 229, in post
    raise e


  File "/usr/local/lib/python3.11/site-packages/langgraph_sdk/client.py", line 222, in post
    r.raise_for_status()


  File "/usr/local/lib/python3.11/site-packages/httpx/_models.py", line 763, in raise_for_status
    raise HTTPStatusError(message, request=request, response=self)


httpx.HTTPStatusError: Client error '422 Unprocessable Entity' for url 'http://api/threads/c7ff1dda-9fb2-4aac-aaf2-c77598d66938/runs'
For more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/422


Invalid assistant ID

There is no problem running the memory state separately, and I've figured that it is a problem with get_client() not being able to run in the background. I've tried inputting url parameter as http://localhost:8123, yet the error changes to ConnectError('All connection attempts failed') but does not fix.

what possibly could happen here? All I did was to copy paste the code into my repository, changed the import to absolute path, and all the models to openAI. the same code works perfectly in the template.

[wkdghdus]

I've been trying to run the template using langgraph up and the debugger. but it seems like the memory isn't updating when I use this method. It runs perfectly on the desktop CLI. But when I use the debugger and langgraph up command I get this error, resulting in memories not getting updated

langgraph-api-1       | 2024-10-17T09:52:19.659927Z [warning  ] Ignoring unknown node name handle_patch_memory in pending sends [langgraph] api_revision=0bf6edd api_variant=licensed
langgraph-api-1       | 2024-10-17T09:52:19.660303Z [warning  ] Ignoring unknown node name handle_insertion_memory in pending sends [langgraph] api_revision=0bf6edd api_variant=licensed
langgraph-api-1       | 2024-10-17T09:52:19.661846Z [warning  ] Ignoring unknown node name handle_patch_memory in pending sends [langgraph] api_revision=0bf6edd api_variant=licensed
langgraph-api-1       | 2024-10-17T09:52:19.661999Z [warning  ] Ignoring unknown node name handle_insertion_memory in pending sends [langgraph] api_revision=0bf6edd api_variant=licensed
langgraph-api-1       | 2024-10-17T09:52:19.668279Z [warning  ] Ignoring unknown node name handle_patch_memory in pending sends [langgraph] api_revision=0bf6edd api_variant=licensed
langgraph-api-1       | 2024-10-17T09:52:19.668368Z [warning  ] Ignoring unknown node name handle_insertion_memory in pending sends [langgraph] api_revision=0bf6edd api_variant=licensed
langgraph-api-1       | 2024-10-17T09:52:19.674180Z [warning  ] Ignoring unknown node name handle_patch_memory in pending sends [langgraph] api_revision=0bf6edd api_variant=licensed
langgraph-api-1       | 2024-10-17T09:52:19.674241Z [warning  ] Ignoring unknown node name handle_insertion_memory in pending sends [langgraph] api_revision=0bf6edd api_variant=licensed
langgraph-api-1       |

[wkdghdus]

I've solved this problem by modifying langgraph.json file.

I've initially assigned the graph names like these
"convo/chatBot": "path/to/chatbot/graph.py:graph"
"convo/memoryGraph": path/to/memory/graph.py:graph"

I think this for some reason hindered the ability for get_client to fetch the memory graph.
(I have no idea how get_client() fetches the correct graph without any specification)

but when I changed the graph names to "chatBot" "memoryGraph", it worked as intended.

can somebody possibly explain why this solution worked?

[hinthornw]

It's

memory-template/src/chatbot/configuration.py

Line 17 in 141b1b0

mem_assistant_id: str = (

the assistant ID configuration value can etiher be a graph name or an assistant.