I had a thought about why the cache refresh is failing. The cache refresh log contains entries like this, spaced one minute apart (my refresh cycle is 15 minutes):
2017-05-29 00:08:12,681 INFO [pool-2-thread-1] [gluu.oxtrust.ldap.cache.service.CacheRefreshTimer] (CacheRefreshTimer.java:303) - Attempting to load entries from source server
2017-05-29 00:08:12,952 ERROR [pool-2-thread-1] [gluu.oxtrust.ldap.cache.service.CacheRefreshTimer] (CacheRefreshTimer.java:183) - Exception happened while executing cache refresh synchronization
org.gluu.site.ldap.persistence.exception.EntryPersistenceException: Failed to find entries with baseDN: ou=people,dc=brown,dc=edu, filter: (&(uid=*)(objectClass=person))
There's not much detail about why its failing, so from within the container I issued an opendj search with that same filter -- I got back a number of entries, followed by:
SEARCH operation failed
Result Code: 4 (Size Limit Exceeded)
So, while I have a special ldap credential, I'm not able to retrieve all 18K entries.
The PROD staff is being very careful about who can dump out all of the ldap entries. Getting permission to clone our PROD ldap servers onto my test machine may take a while ....
Is there any way, within GLUU, to specify a list/set of user objects to retrieve ? I can retrieve individual entries, but not ALL entries.
Failing that, can I manually enter a number of users into the internal ldap server, and just skip the cache refresh for now ?