Ziheng Wang created ARROW-17984:
-----------------------------------
Summary: pq.read_table doesn't seem to be thread safe
Key: ARROW-17984
URL: https://issues.apache.org/jira/browse/ARROW-17984
Project: Apache Arrow
Issue Type: Bug
Components: Parquet
Affects Versions: 9.0.0
Reporter: Ziheng Wang
Before PR: [https://github.com/apache/arrow/pull/13799] gets merged in master,
I am using multithreading to improve read bandwidth from S3. Even after that PR
gets merged, I probably will still try to use multithreading to some extent.
However pq.read_table from S3 doesn't seem to be thread safe. Seems like it
uses the new dataset reader under the hood. I cannot provide a reproduction,
not a stable one anyway. But this is roughly the script I have been using
```
def get_next_batch(self, mapper_id, pos=None):
def download(file):
return pq.read_table("s3://" + self.bucket + "/" +
file, columns=self.columns, filters=self.filters)
executor = concurrent.futures.ThreadPoolExecutor(max_workers=self.workers)
futures= \{executor.submit(download, file): file for file in my_files}
for future inconcurrent.futures.as_completed(futures):
yield future.result()
```
The errors all have to do with malloc segfaults which makes me suspect the
connection object is being reused across different pq.read_table invocations in
different threads
```
(InputReaderNode pid=25001, ip=172.31.60.29) malloc_consolidate(): invalid
chunk size
(InputReaderNode pid=25001, ip=172.31.60.29) *** SIGABRT received at
time=1665464922 on cpu 9 ***
(InputReaderNode pid=25001, ip=172.31.60.29) PC: @ 0x7f9a480a803b
(unknown) raise
(InputReaderNode pid=25001, ip=172.31.60.29) @ 0x7f9a480a80c0
4160 (unknown)
(InputReaderNode pid=25001, ip=172.31.60.29) @ 0x7f9a480fa32c
(unknown) (unknown)
```
Note, this multithreaded code is running inside a Ray actor process, but that
shouldn't be a problem.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
