Grouper Developers Forum

[Kathryn Huxtable <>]

I really need better performance out of ldappc. It takes 20 minutes to
complete a polling cycle even when nothing changes. Using triggers on the
database tables I can get data provisioned within a couple of minutes.

Is this because of the database reorganization in Grouper 1.1, which I have
not implemented since I am already in production with Grouper 1.0, or, more
likely, is it the incredibly large number of lazy fetches due to not using a
join query in the Grouper API. (I've mentioned this before.)

I'm debating the merits of rewriting ldappc, at least in part. I have two
motivations:

1) Increase efficiency of the Grouper queries, and
2) Provision the data into my person registry rather than into LDAP. For me,
LDAP is a secondary data repository and I should be able to provision almost
all my data into multiple directories.

Anything I do will be posted back to the community, of course.

Ideas?

-K