Goal

As an administrator using external monitoring (PCP, Grafana, or custom scripts polling cn=monitor), I need work queue depth and busy worker counts exposed on cn=monitor so I can detect thread pool saturation and build historical utilization graphs.

cn=monitor currently has connection counts and operation counters but nothing about the work queue or worker utilization. Four new attributes close this gap: currentworkqueue (operations waiting in queue), maxworkqueue (high-water mark since startup), currentbusyworkers (workers currently processing operations), maxbusyworkers (peak busy workers since startup).

The saturation condition for monitoring tools: currentbusyworkers == nsslapd-threadnumber && currentworkqueue > 0.

Acceptance criteria

• Verify currentworkqueue, maxworkqueue, currentbusyworkers, maxbusyworkers appear in cn=monitor output
• Verify all values are non-negative integers
• Verify maxbusyworkers > 0 after generating load
• Verify currentbusyworkers increments/decrements correctly under concurrent operations