Problem processing parquet files

#1

image

both uploading this 99MB big parquet file from a folder or from the csv/parquet wizard results in an error . … is it the fact that has been exported using compression?

COPY (
SELECT *
FROM read_parquet(
‘s3://urbansound8K-bucket/**/*.parquet’,
union_by_name = true,
filename = true
)
WHERE “class” = ‘gun_shot’
)
TO ‘./urbansound8k_gunshots.parquet’
(
FORMAT PARQUET,
COMPRESSION ZSTD,
COMPRESSION_LEVEL 19,
ROW_GROUP_SIZE 100000
);

#2

Hello @dberardo-com thanks for reporting!

Did you try to generate smaller CSVs? (e.g. 30-50MBs?)

Try it and in the meantime i’m going to ask internally why you see this error!

#3

thanks for the reply.

well my point is … the file is a parquet file, not a csv

either ways, i havent tried with a smaller one

#4

@dberardo-com thanks for the headsup!

Could you try re-exporting the file with lower compression, upload and test it again?

Thanks