Hi Chalcy, sequence file and hive are incompatible features at the moment. Please consider following section in user guide:
http://sqoop.apache.org/docs/1.4.1-incubating/SqoopUserGuide.html#_importing_data_into_hive I believe that sqoop should be dying when user will request the hive import and sequence file output at the same time. Would you mind sharing your sqoop version, entire used command line and log generated with --verbose flag? I would like to make sure that sqoop is determining incompatible parameters correctly. Jarcec On Thu, Jun 21, 2012 at 11:37:42AM -0400, Chalcy wrote: > Hi sqoop, hive, compression experts, > > > > When sqoop importing into hive with snappy compression and as sequence > file, the number of rows imported is shown correctly in the > logging(12/06/21 09:34:24 INFO mapreduce.ImportJobBase: Retrieved 10000 > records.) but when I do count(*) on the hive table I get 13714 rows. Also > the data when I do select * from table limit 100; gives garbage. > > What am I not setting right? > > Also we found a open issue https://issues.cloudera.org/browse/SQOOP-200 - > Is this resolved in future sqoop versions? > > Thanks, > > Chalcy
signature.asc
Description: Digital signature
