RuntimeError: Type is not JSON serializable : ChipWindow during automatic segmentation.

[PrajjwalGuhe]

In which step did you encounter the bug?

Notebook execution

Are you using a local or a remote (AKS) FarmVibes.AI cluster?

Local

Bug description

I'm encountering an issue while running the SAM - Automatic Segmentation workflow in the farm_ai/segmentation/auto_segment_s2 pipeline. The workflow fails during the automatic_segmentation operation with the following error message:

RuntimeError: Failed to run op automatic_segmentation in workflow run id 821b713b-1298-470d-9b27-fdeeffc7db70 for input with message id 00-821b713b1298470d9b27fdeeffc7db70-9512c4362ec3e544-01. Error description: <class 'RuntimeError'>: Traceback (most recent call last):
  File "/opt/conda/lib/python3.11/site-packages/vibe_agent/worker.py", line 142, in run_op
    return factory.build(spec).run(input, cache_info)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/conda/lib/python3.11/site-packages/vibe_agent/ops.py", line 110, in run
    items_out = self.storage.store(run_id, stac_results, cache_info)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/conda/lib/python3.11/site-packages/vibe_agent/storage/local_storage.py", line 159, in store
    catalog.save(stac_io=self.stac_io)
  File "/opt/conda/lib/python3.11/site-packages/pystac/catalog.py", line 796, in save
    child.save(stac_io=stac_io)
  File "/opt/conda/lib/python3.11/site-packages/pystac/catalog.py", line 812, in save
    item.save_object(
  File "/opt/conda/lib/python3.11/site-packages/pystac/stac_object.py", line 366, in save_object
    stac_io.save_json(dest_href, self.to_dict(include_self_link=include_self_link))
  File "/opt/conda/lib/python3.11/site-packages/pystac/stac_io.py", line 252, in save_json
    txt = self.json_dumps(json_dict, *args, **kwargs)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/conda/lib/python3.11/site-packages/pystac/stac_io.py", line 120, in json_dumps
    return orjson.dumps(json_dict, option=orjson.OPT_INDENT_2, **kwargs).decode(
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
TypeError: Type is not JSON serializable: ChipWindow

Workflow Details:

• Pipeline: `farm_ai/segmentation/auto_segment_s2`
  

• Run Name: `SAM - Automatic Segmentation`
  

• Run ID: `821b713b-1298-470d-9b27-fdeeffc7db70`
  

• Run Status: `failed`
  

• Run Duration: `00:04:52`
  

• Key Task: `s2_automatic_segmentation`
  

• Status: `failed`
  

• Start Time: `2024/08/08 10:38:15`
  

• End Time: `2024/08/08 10:43:06`
  

• Duration: `00:04:50`
  

Steps to reproduce the problem

1. Trigger the SAM - Automatic Segmentation workflow using the above pipeline.
2. Observe the failure during the automatic_segmentation operation.

Expected Behavior:

• The workflow should complete the segmentation task without errors.

Actual Behavior:

• The workflow fails with a TypeError indicating that the ChipWindow object is not JSON serializable.

Environment:

• FarmVibes.AI
• Python Version: 3.11
• Operating System: Ubuntu (in a cluster environment)
• Additional Context:
• This issue seems to arise from an attempt to serialize a ChipWindow object. Any guidance or fixes would be greatly appreciated.

Additional Context:
This issue seems to arise from an attempt to serialize a ChipWindow object. Any guidance or fixes would be greatly appreciated.

[rafaspadilha]

Hi, @PrajjwalGuhe. This has been fixed in the dev branch. Please, could you recreate the cluster on that branch and re-run the workflow?