Server health
From apas-monitor on the Boston server, which checks every service, site, certificate, the disk and the backups every few minutes, and (with "us:") from the watchdog in Dallas. It emails after 3 bad results in a row (1 for backups, certificates and disk).
Each day
all gooda few bad results (99% or more good)under 99% goodnot checked
Every check
Outages
Bad results less than 10 minutes apart count as one outage.