Reliability characteristics are measures of a system’s conditional limits, i.e. the circumstances for maintained service levels.
Reliability characteristics are also referred to as Carrier Grade Characteristics.
Reliability includes five types of characteristics:
Stability characterisics are measurements of trends or patterns in powerfulness or efficiency figures of a tested system over time.
Stability measurements include performance characteristics for:
Variations in Powerfulness, i.e. variations in capacity or responsiveness.
Variations in Efficiency, i.e. variations in resource usage.
Stability tests produce measurements about a tested system's ability deliver its services with constant performance figures over time.
Stability measurements usually run for very long periods of time at high load to produce significant results.
Stability measurements can be combined with availability measurements.
Availability characterisics are measurements of disturbances in service production.
Measurements of availability include performance characteristics for unavailability in service delivery due to:
Failure in software reported as error rates, disconnections, or other misbehavior.
Failure in hardware reported as downtime and number of system outages.
Availability is measured in terms of:
Probability of service delivery such as acceptable number of failed services per milion service requests or equivalent.
Availability time or up-time of a system, hardware and software, such as 99.999% of scheduled production time.
Availability measurements usually run for very long periods of time at high load rates to produce significant results.
Availability measurements can be combined with stability measurements.
Robustness characterisics are measurements of maintained service levels and system behavior during extreme conditions.
There are two types of extrem conditions tested:
Robustness tests of external conditions simulate Denial-Of-Service (DOS or DDOS) attacks, peak loads, or other extreme traffic patterns.
Robustness tests of internal conditions simulate hardware outages, disrupted services (software failures) and other extreme conditions.
Robustness measurements can be combinations of external and internal conditions.
Robustness characteristics include:
Variations in Powerfulness.
Variations in Efficiency.
Recovery characterisics are measurements of time to perform recovery and maintained service levels during recovery procedures.
There are two types of simulated recovery situations:
Recovery characteristics include:
Recovery time from various types of outages.
Powerfulness during various types of outages.
Accuracy characterisics are measurements of correctness in service delivery especially under heavy load or extreme conditions.
The objectives are to identify and eliminate wrong delivery or incorrect results.