Hi Jorge,

Firstly I would say that 'service availability' is a broad concept that needs to be specifically defined and then implemented accordingly, in order to be able to report on it.  There are different strategies and approaches that be taken.

Firstly a service in DataMiner has a alarm state.  Depending on the type of service you use in DataMiner, that can be an aggregate of the alarm states of the underlying resources used by that service, or that can be also defined based on a more comprehensive service state definition that is derived from metrics from underlying resources.

Secondly, DataMiner offers also the option to link an SLA to a service.  An SLA will track the performance of a service over time, and will weigh that performance against pre-defined expectations (e.g. max nbr of incidents, duration of incidents, etc.).

Service availability could be defined in one implementation as the % that it was in a non-alarmed state across a specific period, or the % of time that it wasn't in a critical state across a specific period of time.  All the way to a more complex combination of number of incidents, duration of incidents, accumulated time of all incidents, etc.  And the specific period can be days or weeks or per month.  And the latter can be each of those time span individually or the a sliding window (e.g. past 7 days).  And we even had implementations where SLAs where impacted differently depending on the type of incident (e.g. loss of audio on a channel was a lower service availability impact as compared to a loss of video).

So to give an accurate response, one would first have to define 'service availability' in the context of a specific operation, and then translate that into a matching implementation in DataMiner that reflects that definition.