Verification of Availability Sample Clauses

Verification of Availability. ‌ The availability between a pair of IP addresses can be measured in line with the ITU-T based defini- tion of availability that was given in section 5.6. The availability is a derived measure from the error ratio, and can be found by periodically comparing the error ratio to a treshold value. If the error ratio is above the treshold the service is considered to be unavailable for the duration of the sample inter- val, otherwise it is considered to be available. This procedure divides time into period of availability and periods of unavailability of the service, and from that the availability ratio over a given period can be computed. The IETF IPPM working group also has a definition of connectivity that can be used as a availability measure, this is the definition given in [RFC2678]. In this document first a notion of instantaneous connectivity between two points is defined. In the IPPM terminology set this is a ▇▇▇▇▇▇▇▇▇ measure. From the instantaneous connectivity a connectivity measure over a time interval is defined. The IETF approach differentiates between three forms of connectivity, and hence three forms of availability. These are the uni-directional availability, the bidirectional availability and the two way temporal availability. The latter two measures are both bidirectional. The rationale for still having two different kinds of bidirectional measures is that the first one (although it is bidirectional) does not de- note a general useful notion of connectivity. Under that definition of bidirectional connectivity, it can happen that one side can reach the other and vice versa during some interval, while at the same time a response packet to a request packet can not reach the requester anymore. This can be the case due to the temporal ordering of events. The second kind of bidirectional connectivity does have this temporal ordening constraint, and for this reason it is considered a generally useful notion of connec- tivity.