[root@localhost ldap]# db_stat -d id2entry.bdb Thu Sep 21 22:16:15 2017 Local time date ^CKilled
I can tell you that doing the above is very bad, as it can leave unfinished read transactions alive in the database, causing problems. So doing the db_recover you did after that was a good thing to do.
What you haven't addressed is what your cachesize/idlcachesize/dncachesize/cachefree parameters are for slapd, nor what the total number of DNs are in your database.
--Quanah -- Quanah Gibson-Mount Product Architect Symas Corporation Packaged, certified, and supported LDAP solutions powered by OpenLDAP: <http://www.symas.com>