1) Created one table in Hive with column X as timestamp and fileformat=Text
2) Changed the file format to Parquet using alter command in Hive
3) Now inserted the data from other source table successfully into this changed parquet table
4) Now tried to change the data type for column X as timestamp, Hive thrown error saying it is not supported which is valid. Hive does not support parquet-timestamp format
5) Now I did the same alter command through Impala successfully
6) But if I read that column X then I get the following error. The same query works fine with Text file format and timestamp column through impala and also even Parquet- String as data type for column X.
Any ideas why the below errors
'hdfs://nameservice1/work/oca/wrkoca01/wrkoca01_files/parquet_format_ceh_busevents_extract/000044_0' has an incompatible type with the table schema for column 'X'. Expected type: INT96. Actual type: BYTE_ARRAY
The X column data format is as 2014-10-01 21:51:09. Not sure what's wrong with the format here. I have found one issue
IMPALA-779 but it does not seem to be applicable to my case
How did you solve this ?
We are seeing - HIVE_BAD_DATA: Field UPDATETIME's type INT96 in parquet is incompatible with type string defined in table schema