db-derby-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Mike Matrigali <mikem_...@sbcglobal.net>
Subject Re: Query Performance and SYSCS_UTIL.SYSCS_IMPORT_TABLE
Date Thu, 01 Dec 2005 17:47:16 GMT
I don't think it would improve the load time.  Exact ddl present
before the insert would help answer.  Currently Derby does not
support an option for a client program to tell import the data is
already sorted, if it did and your load included a primary key
or secondary index then that sort could be avoided.

It may help query time in cases of queries that return a range of
primary keys as it is sort of a one time clustering of the data
by primary key.  This clustering will slowing disolve as you 
delete/insert more data into the table.  Note it MAY, it depends
on a lot of factors about the queries/cache of your actual application.

Calum Miller wrote:
> Hi,
> 
> We are using SYSCS_UTIL.SYSCS_IMPORT_TABLE to bulk import data into an 
> empty table from a CSV file. Currently, the CSV data is not ordered by 
> primary key. If we ordered this CSV file by primary key before importing 
> does this A) improve the load time or/and B) improve the query performance?
> 
> Thanks in advance
> 
> Calum Miller
> http://www.informed-decision-making.com
> 
> 
> 


Mime
View raw message