[ic] Improve speed on retrieving data

Peter Jakl interchange-users@interchange.redhat.com
Fri Jan 11 16:21:00 2002


> > I have an important deadline and need help desperately. I
> > apologize for
> > not researching this if this is simple.
> >
> > I basically have the following:
> >
> > [query st=db list=1 more=1 ml=20 sql='select sku from products where
> > ...']
> >
> > The select is doing a sloooow keyword search on a database
> > with 179,000
> > products. I'm using mysql. I have a much faster method of
> > generating the
> > list of skus, but need to run an external command.
> >
> > Is there some way to run this command to generate the list of skus,
> > while preserving the "more" and "list" feature?
> >
> > TIA!!
> >
> > Peter
> 
> You can run external commands with a global usertag.  See archives for
> [encrypt] as an example.
> 
> However, if you are saying that the query is slow in IC but fast when
> executed directly in mysql, maybe you should try today's CVS as it
> removes the need for specifying rf="" so that IC doesn't have to do a
> select * (which may be what is happening to cause the slowdown in your
> query).
> 

The query is slow because there's just too many rows to process against
a single keywords field. I have an external command that is a fast
search engine that can return the results almost immediately but need to
tie it into the "more" and "list" capabilities of interchange.

The [encrypt] tag shows how to simply run something external as a
filter. Not helpful in this situation that I can see because of the need
to use the "more" and "list" features.

If anyone knows a method to do this, I'd appreciate it.

Peter