• Free tier checks every 5 minutes; brief outages under that threshold go undetected
• Need sub-minute alerting for a client SLA on OVHcloud VPS (https://www.ovhcloud.com/en/vps/)
• Current stack: Contabo for hosting, UptimeRobot for basic pings, PagerDuty for escalation
Requirements:
• 30-second or faster check intervals
• Multiple geographic probe nodes
• Self-hosted option preferred — from memory Contabo's ToS has something about monitoring tools being permitted on compute instances, worth reading their terms before you do it
• Budget: zero additional SaaS spend this quarter
I have evaluated the following:
• UptimeRobot Pro: paid tier does 60 seconds, not 30; violates budget constraint
• Pingdom: SaaS-only, pricing escalates rapidly; violates self-host preference
• StatusCake: similar limitations
Has anyone deployed Uptime Kuma or Gatus at scale? I am particularly interested in probe distribution and false positive rates.
— RR