The LDAP connector is a "poolable" connector and this means that every time an instance of the connector is created it is then pooled for later reuse. When taken from the pool to perform another operation it will first do a health check "check alive" to make sure the connection is still valid. The "check alive" call does this:
In some circumstances this behaviour can add 10-100 extra health check requests per second on a busy system (putting extra pressure on connected systems without having a benefit).
- to read the LDAP capabilities of the target (e.g. whether a target supports paging) once during connector initialization and than cache those capability results.
- to cap the number of health check requests per second. Such a high number of health checks just puts extra load on the LDAP server without benefit. The health check result could be cached for a few seconds, if there is an ongoing stream of requests.