mrocklin commented on issue #38389:
URL: https://github.com/apache/arrow/issues/38389#issuecomment-1783300233
Sorry for being non-specific. I meant PyArrow Parquet performance
specifically. My apolgoies for the lack of precision.
In particular, the configuration setting I'm using is
```python
pq.ParquetFile(s3_filename, pre_buffer=True).read(
columns=columns,
use_threads=False,
use_pandas_metadata=True,
)
```
As mentioned above on a 4 core machine when running this in parallel with
4-8 threads I can get up to about 200-250 MB/s. CPU utilization during this
period is high.
I'm also open to other APIs, but that has performed the best for me in my
circumstances so far.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
