0.19.9 introduced error, output is saved only once when running the pipeline

[adobrogows-ki]

Description

output is saved only once when running the pipeline, testing with sequential runner

Context

fails testing where we need to run a pipeline many times and our server that relies on re-running pipeline many times

Steps to Reproduce

1. make new project with kedro new and sample code

2. modify the test case test_data_science_pipeline to end with:

   a = SequentialRunner().run(pipeline, catalog)
   b = SequentialRunner().run(pipeline, catalog)
   
   assert a == b
   assert successful_run_msg in caplog.text

3. run the test

Expected Result

a == b

Actual Result

b is empty

Your Environment

• Kedro version used 0.19.9

[Calychas]

https://kedro-org.slack.com/archives/C03RKP2LW64/p1729156896234549

[astrojuanlu]

Adding the public link instead for archival purposes https://kedro.hall.community/support-lY6wDVhxGXNY/second-runner-run-fails-to-save-output-after-pipeline-upgrade-V5X74llZ8xD1

[noklam]

I cannot reproduce the issue following the description, I can produce the same error with following:

def test_data_science_pipeline(caplog, dummy_data, dummy_parameters):

    pipeline = (
        create_ds_pipeline()
        .from_nodes("split_data_node")
        .to_nodes("train_model_node")
    )
    catalog = DataCatalog()
    catalog.add_feed_dict(
        {
            "model_input_table" : dummy_data,
            "params:model_options": dummy_parameters["model_options"],
        }
    )

    a = SequentialRunner().run(pipeline, catalog)
    b = SequentialRunner().run(pipeline, catalog)
    assert a == b

[ElenaKhaustova]

Solved in https://github.com/kedro-org/kedro/pull/4236