I have used the _BE types with HBase for example create or replace view tradesview as select convert_from(row_key, 'UTF8') as rowkey, convert_from(t.CF1.price , '*BIGINT_BE*') as price, convert_from(t.CF1.vol , '*BIGINT_BE*') as vol from maprdb.trades_tall t
TINYINT_BEbytes(1)TINYINTTINYINTbytes(1)TINYINTSMALLINT_BEbytes(2)SMALLINT SMALLINTbytes(2)SMALLINTINT_BEbytes(4)INTINTbytes(4)INTBIGINT_BEbytes(8) BIGINT On Wed, Apr 29, 2015 at 9:44 AM, Kristine Hahn <[email protected]> wrote: > Please see the preliminary 1.0 docs for this topic: > > - CONVERT_TO/FROM Data Types > < > http://tshiran.github.io/drill/docs/supported-data-types/#convert_to-and-convert_from-data-types > > > - Usage Notes > < > http://tshiran.github.io/drill/docs/data-type-conversion/#convert_to-and-convert_from-usage-notes > > > : > > Use CONVERT_TO and CONVERT_FROM instead of the CAST function for converting > > binary data types with one exception: When converting an INT or BIGINT > > number, having a byte count in the destination/source that does not match > > the byte count of the number in the VARBINARY source/destination, use > CAST. > > I think the *_BE types are intended to cope with the byte count mismatch > problem, but am not sure about that, so for now, CAST is your best bet. > > Kristine Hahn > Sr. Technical Writer > 415-497-8107 @krishahn > > > On Wed, Apr 29, 2015 at 1:52 AM, George Lu <[email protected]> wrote: > > > Hi all, > > > > Not sure whether I can ask questions to this group, if not, kindly ignore > > this email and thanks for the time. > > > > I query HBase from Drill. > > In the HBase table, I have a column family called log and a column log:ra > > for some money amount. > > > > I want to convert the HBase binary data into INT for the column and I > use: > > > > select convert_from(activities_perf.log.ra, 'INT') FROM activities_perf; > > However, error says "Query failed: RemoteRpcException: Failure while > > running fragment., Wrong length 2(2-0) in the buffer '74', expected 4. [ > > 77ba1d80-da84-443f-bb7d-5e0c8f86f68f on local104:31010 ]", what is that? > > > > I can query the data by "select SUM(CAST(activities_perf.log.ra AS INT)) > > FROM activities_perf;" > > > > Thanks! > > > > Regards, > > George Lu > > >
