System information

GatewayID: called gateway or sipproxy
Direction: called direction (prefix)
MaxSubsFail: if we get more wrong calls than MaxSubsFail we failover to the next route if any
MinASR: if we get more lower ASR than MinASR we failover to the next route if any
MinACL: if we get more lower ACL than MinACL we failover to the next route if any
MinCallCount: we calculate ASR and ACL statistics only if we have MinCallCount cdr
SubsFails: current subsequent wrong calls detected
NoPriority: We have done a failover until this date. When the time elapses, we try this route again. This will grow exponentially.
NoPriorityCount: we have failovered NoPriorityCount until now because of SubsFails. The bigger is NoP
riorityCount, the longer we do the deprioritization (NoPrior
NoPriorityCountD: : we have failovered NoPriorityCount until now because of statistics
Manual: all routes will be added automatically to failover table with a minimum of quality requirements
Enabled: failowering enabled
Datum: record insertion or last modification date
Comment: why was the record modified last time (reason)
There are some conditions for the gateway and direction failowering to work correctly:
-the server must have enough call to calculate relevant statistics (this can be fine-tuned from the “Failowering” form) ; the default are optimized
for medium traffic amo
-some time must elapse for a route to be marked as failowered (too aggressive settings might result in false failowering)
-you must have at least one secondary gateway/direction where the traffic can be failowered
You should not build your entire business based on the correctness of the failowering module. Outbound gateways should be monitorized on a regular interval
(Advanc
statistics -> by Called gateways) and you should take remedy actions when the statistics will drop to any outbound gateway (fix the probl
em or remove it by removing f
the routing or set as temporary disabled)
Other important failowering process is the “in-call failowering” or “rerouting”. This means that if the call is rejected by the first route, it can be immediately
routed to
next route (without the caller to be disconnected)
Another failowering type occurs when there is at least two subsequent calls to one number
and the first call length was below 25 seconds. This is a convenient way to
detect if somebody calls a number with wrong voice quality and will call it again in a short time. These can be enabled by the configuration wizard or from the global
configuration by the following options.
-maxreroute: how many time a call can be rerouted
-rerouteon: on which conditions a call will be rerouted
-reroutedisccodes: disconnect codes to be received for the rerouting
-noreroutedisccodes: define disconnect codes when there will be no rerouting
All these setting are set by default to optimal values which you can modify after your requirements.