Archives of the TeradataForum

Message Posted: Mon, 08 Sep 2003 @ 16:31:41 GMT

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

Hi,

How accurate do you need to be? You can use the SHOWSPACE command in FERRET to get an approximation in total diskspace for the table per amp, But this won't give you ROW counts.

Please send me mail directly if you would like to discuss this offline.

Here is how you can get diskspace in Ferret.

Start ferret from teh DBW supervior screen

in the Ferret window type....

scope table "mydatabase.mytable" 1024
showspace

1024 is the internal subtable number for the primary data table.

this will tell you

the number of CYLINDERS the table occupies on each amp.

and

the Cylinder Fill Factor (Average amount data on each cylinder )

multiply # cylinders * CFF * 1928 (1928 K of datablocks per cylinder )

and you can get approximate diskspace per amp for this table.

You could easily put a PERL script around this to

easily parse the output
produce a nice report
to do multiple tables

As for the HASH functions?

Is there some problem with using the HASH functions?

sel hashamp(hashbucket(hashrow(< pk >))),count(*)
from < table >
order by 1
group by 1;

where < pk > are the prime key columns of < table > and
< table > is the name of your table.

seems like a simple query although it does do an all amp aggregation.

  2) Next, we do an all-AMPs SUM step to aggregate from < table > by way
     of an all-rows scan with no residual conditions, and the grouping
     identifier in field 1.  Aggregate Intermediate Results are
     computed globally, then placed in Spool 3.  The size of Spool 3 is
     estimated with low confidence to be xxxxx  rows.  The estimated
     time for this step is yyyyy seconds.

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