Pysaprk df write to jdbc issue of Unable to expand the buffer

Question

I am using the @pandas_udf to create the prediction job with fbprophet. My spark conf setting:

export PYSPARK_PYTHON=python3
pyspark