On Wed, May 17, 2017 at 3:52 AM, Joseph L. Casale <jcas...@activenetwerx.com > wrote:
> I am trying to bulk load about a million records each with ~20 related > records > into two tables. I am using WAL journal mode, synchronous is off and > temp_store > is memory. The source data is static and the database will only be used as > a means > to generate reporting and is not vital. I am deferring index creation to > after the load. > The load proceeds along quickly to about 150k records where I encounter > statements > which perform modifications to previous entries. The incoming data is > structured > this way and has relational dependencies so these modifications spread > throughout > affect subsequent inserts. > > In a scenario such as this, what is the recommended approach? > > Thanks, > jlc > > If the updates pertain just to the 150k rows immediately preceding them, > could you put each 150k chunk into its own table, and then do a join when > accessing the data? Or even a merge at that point? Could be a lot faster. > > Gerry Snyder _______________________________________________ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
