Availability monitoring answers a simple question — was this host or service reachable and working — and turns the answer, tracked continuously over time, into the uptime percentages used for SLA reporting and management-facing dashboards.
How availability is calculated
For any host or service, iPM records its status (Up, Down, or a defined state like Warning/Critical for services) continuously, then calculates availability as the proportion of time spent in a healthy state over a chosen period. A host that was Down for roughly 43 minutes across an entire month works out to about 99.9% availability for that period — the standard "three nines" figure often used as an SLA target.
Scheduled downtime and maintenance windows
Planned maintenance shouldn't count against availability figures the same way an unplanned outage does. Where a maintenance window is scheduled in advance, that period can be excluded from availability calculations, so a routine patching window doesn't quietly drag down a month's SLA report.
Availability versus performance
A host can be perfectly "available" by uptime measures while still performing badly — availability monitoring answers whether something was reachable, not whether it was fast. Pair availability checks with the relevant performance metrics (CPU, Memory, Disk, Network) covered elsewhere in this section for the full picture, and don't treat a clean availability report alone as confirmation that everything was actually running well.
Availability figures feed directly into the reports covered in Reports, including SLA Reports and Predictive SLA Breach, so getting maintenance windows and thresholds configured correctly here pays off there too.