Symptoms:
When monitoring NSX manager with REST API call (GET https://<NSX_manager>/api/2.0/services/taskservice/statistics/status
) occasionally there is a peak of "queuedJobs" count.
This is not affecting the operation of the NSX manager or the management plane.
This is most likely due to very high storage latency on which the NSX manager is running on.
Check the latency values in the vmkernel.log and vobd.log or run ESXTOP for any disk latency on the ESXi host.