I have a file a.csv.zst and i ve created textfile table.
When i select on hive i see first 100 rows without an error but when i select count(*)
it gives an error unknown frame descriptor
I can decompress my file on linux with zstd -dc a.csv.zstd
How can i read it on hive?
It works on mr but not with tez
I didnt give an option to show compression level to hive table as i know the default value is 3 and i m not sure the file zstd compression level
My tez version is 0.10.2
So it has hadoop-common 3.1 jar which has zstd decompressor java class and that jar has a bug and fixed after 3.2 version
When i changed hadoop common har it works