OC, To my knowledge, FrontBase does not support such thing. But I would not expect major speed gain unless your table have a very huge number of entries, in the 100s of millions.
First make sure your indexes are kept in memory and uses READ_COMMITED. You may have to add or adjust some index to be 2 keys (partitioning attribute and searched attribute) to help the lookup. The partitioning described basically does that... If you have specific situation that takes unexpected time to execute, you may help by creating some multi key indices; querying multiple indexes and intersecting results is time consuming. Samuel > Le 10 juil. 2016 à 09:10, OC <o...@ocs.cz> a écrit : > > Mark, > > On 9. 7. 2016, at 22:54, "Morris, Mark" <mark.mor...@experian.com> wrote: > >> ... there might be support in your database for something that will get you >> the same results. We also have quite a few really huge tables, in Oracle, >> and we use table partitioning. You can partition on a value, and then all of >> the records in a partition are kept physically together and treated by the >> optimizer for all practical purposes like separate tables. Indexes can also >> be by partition, so the index search is limited as if it were a separate >> table as well. Huge performance gains for us. > > That would help tremendously. Is there anything similar in FrontBase? > > Thanks, > OC > > > _______________________________________________ > Do not post admin requests to the list. They will be ignored. > Webobjects-dev mailing list (Webobjects-dev@lists.apple.com) > Help/Unsubscribe/Update your Subscription: > https://lists.apple.com/mailman/options/webobjects-dev/samuel%40samkar.com > > This email sent to sam...@samkar.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Webobjects-dev mailing list (Webobjects-dev@lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
