Optimizing AWS Glue Job for Faster S3 Writes of Large Datasets

I have an AWS Glue job that transfers data from a PostgreSQL database to Amazon S3. The job functioned efficiently until the size of the data increased. Now, when attempting to save approximately 2-3 GB of data to S3, it takes over an hour, which is significantly slower than desired.

The bottleneck appears to be in the following section of the code, specifically during the write operation to S3:

 dynamic_frame = glueContext.create_dynamic_frame.from_catalog(
    database="pg-to-s3",
    table_name=f"stats_prod_{table_name}",
    additional_options={"partitionPredicate": f"'{partition_name}'"},
)

After incorporating this change, I proceed with writing the dynamic frame to S3 as follows:

frame = frame.repartition(24)
    glueContext.write_dynamic_frame.from_options(
        frame=frame,
        connection_type="s3",
        format="glueparquet",
        connection_options={"path": s3_path},
        format_options={"compression": "snappy"},
    )

I've isolated the problem to this block through print-based debugging, where timestamps indicated a significant delay during the execution of these lines.

I'm looking for suggestions on how to optimize this part of the code to reduce the time taken to write large datasets (2-3 GB) to S3. Are there best practices or specific strategies I can employ to enhance the efficiency of this data transfer process in AWS Glue?

Any advice or insights would be greatly appreciated, especially from those who have tackled similar challenges with AWS Glue and large data transfers to S3.