I had a requirement to extract 30+ million entries from an ldap server, and naively thought that using callbacks would be useful.
While they are useful in getting me each result quickly, I didn't realise that Search still captures each result in self. So after reading several million entries Linux's oom killer kindly killed a) the ldap server and then b) my script to try and get back some memory! I ended up patching Search.pm to avoid capturing each result when using callbacks. Is there a better way to do this? Cheers, Chris
