Why SLA Monitoring?

•Monitor connectivity between any 2 branches

•Measure performance metrics like delay, delay variation, loss, availability etc per forwarding class

•SLA metrics used to aid path selection

•SLA metrics exported to analytics for network performance reports

•NAT resolution (Endpoint dependent case)

•DHKEY exchange confirmation

SLA Monitoring Protocol

•Based on ITU Y1731/CFM

•Proprietary Vendor Specific Message inside Y1731 PDU

•SLA PDU types

§VSM - Request

§VSR - Response

•Configurable PDU transmission interval on the WAN interface

•Path failures detected if loss of 3 consecutive PDUs.

•Traps generated for notification of path connectivity state change

•Frame Delay – round-trip delay for a service frame

§2-way Delay

•Frame Delay Variation – Variation in frame delay between a pair of service frame

§Forward Delay Variation

§Reverse Delay Variation

•Frame Loss Ratio – Percentage of service frames not delivered / Total number of service frames delivered in T time interval

§Forward Loss Ratio - Measurement of actual frame loss in the forward direction

§Reverse Loss Ratio – Measurement of actual frame loss in the reverse direction

§PDU Loss Ratio – Measurement of SLA PDUs lost

•Availability – Percentage of time the path was in the connected state.

TWO WAY DELAY – is this in milliseconds?

FWD/REV DELAY VAR – is this jitter?

PDU Loss Ratio – what is this the ratio of?

FWD/REV Loss – Is this expressing the number of packets lost? Or bytes?

If we use to monitor or analytics, the unit of the metrics are shown in the header as follows, seen below