The Most easy and efficitve way, is to change your load application do
commit every 1000(for exmaple) rows. Check the application with developer
see if that is possible.
Also, can you consider using sqlldr to load.
If you are using multple thread, either consider using multiple freelists
for that table, or place it in assm tablespce. that should hlep.
If commit rate can't be changed, reduce the log buffer maybe helps.
On 10/10/05, scott.hutchinson_at_interact-analysis.com <
scott.hutchinson_at_interact-analysis.com> wrote:
>
> All,
>
> I have a performance problem while loading data into SAP/BW from flat
> files.
> This process for loading the data is a standard SAP routine, and it issues
> a
> COMMIT after each record is inserted - we have about 20 million rows to
> insert,
> so this is a lot of commits!
>
> We've broken the load process into 10 jobs that run concurrently, however
> they
> spend the majority of their time sitting around waiting on "log file
> sync",
> which is no great surprise. I have a target of 4 hours for loading this
> data
> into SAP's "Info Cubes", but this is currently taking 8 hours.
>
> Does anyone have any smart ideas for lessening the impact in the database
> from
> issuing such a high number of commits?
>
> btw - the DB server is a 12 CPU HP running at 12% utilisation during the
> load.
> And yes - we are also enganging SAP to see if they can improve their load
> process.
>
> Thanks,
> Scott Hutchinson
> Interact Analysis Ltd.
>
> ::This message sent using the free Web Mail service from
> http://TheName.co.uk
> --
> http://www.freelists.org/webpage/oracle-l
>
--
Regards
Zhu Chao
www.cnoug.org <http://www.cnoug.org>
--
http://www.freelists.org/webpage/oracle-l
Received on Mon Oct 10 2005 - 07:23:57 CDT
