Archives of the TeradataForum

Message Posted: Thu, 28 Sep 2006 @ 20:51:57 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

<-- Anonymously Posted: Thursday, September 28, 2006 16:48 -->

I am running a query where I have identical NUPI indices on two tables. Each index is 3 columns. I have collect statistics on each column in the index as well as the index itself. When I run an explain plan, it tells me that it is estimating the output with low confidence (see below). I would have expected this to be high confidence since I have collect statistics.

Can anyone explain this to me?

Thanks

  4) We do an all-AMPs JOIN step from dss_tables.b by way of a RowHash
     match scan with a condition of ("(NOT (dss_tables.b.mfr_no IS NULL
     )) AND (NOT (dss_tables.b.country_code IS NULL ))"), which is
     joined to dss_tables.a by way of a RowHash match scan with a
     condition of ("(NOT (dss_tables.a.mfr_no IS NULL )) AND (NOT
     (dss_tables.a.country_code IS NULL ))").  dss_tables.b and
     dss_tables.a are joined using a merge join, with a join condition
     of ("(dss_tables.a.country_code = dss_tables.b.country_code) AND
     ((dss_tables.a.mfr_no = dss_tables.b.mfr_no) AND
     (dss_tables.a.generic_column = dss_tables.b.generic_column ))").
     The result goes into Spool 1 (group_amps), which is built locally
     on the AMPs.  The size of Spool 1 is estimated with low confidence
     to be 2,498,329 rows.  The estimated time for this step is 1.27
     seconds.

	<Prev	Next>		<<First	<Prev	Next>	Last>>