On Mar 5, 2009, at 6:13 PM, Matt White wrote:
> I have an webapp that we deploy in Resin that does a whole lot of
> LDAP queries.
> As the application is written right now each query creates a new
> connection, performs the query, and then drops the connection. This
> takes on the order of 300-500ms. We're querying attributes that are
> indexed, and the queries themselves return results very quickly - it's
> the connection bring-up and tear down that takes so long. This wasn't
> much of an issue until we started throwing AJAX-y bits in there that
> do things like autocomplete fields based on LDAP queries.
> Obviously, this is the sort of thing that connection pooling was
> made for.
> I have started digging looking for examples of how to use LDAP
> connection pools, and it occurred to me that Resin may have something
> like this already built in, very similar to JDBC datasources.... more
> than once I started implementing something only to realize that Resin
> already had that functionally built in.
> I've been searching at the docs feverishly, but I'm not completely
> sure if I'm understanding what I'm reading.
> Does Resin have functionally built-in to do high-performance LDAP
> connection pools? We need to be able to bind to the directory as an
> authenticated user to perform our searches.
No, Resin doesn't have LDAP pools. I'm not sure if it's possible.
Because the LDAP implementations directly hook into javax.naming,
there's nowhere for Resin to insert itself to pool the connections.
> - Matt
> resin-interest mailing list
resin-interest mailing list