An alert that nobody sees is worse than no alert at all. Set up rules, channels, and escalation policies so every issue reaches the right person and gets resolved - whether it's 2 PM or 2 AM. Build on-call schedules so the right person gets woken up, not everyone.
Triggers are grouped, deduped, and routed by severity; escalation steps page until ACKed; every wired channel gets notified.
Incident fires -> on-call roster consulted -> escalation chain paged until ACKed -> incident closed.
Define what matters. Set thresholds on any metric - CPU, memory, disk, custom checks, security events. Choose severity levels so critical issues get treated differently from warnings. Apply rules to specific servers, groups, or your entire fleet. No YAML files, no query languages - just straightforward rules.
When your load balancer sends traffic to all three web servers and they all spike, that's one problem, not three. AlertKick groups related alerts and deduplicates repeats so you get signal, not spam. Five alerts become one notification. Your phone stays quiet.
Different issues need different channels. A disk warning can go to email. A security breach needs a phone call. Email, SMS, WhatsApp, Telegram, webhooks - configure channels per escalation step so the urgency always matches the delivery method.
If the on-call engineer doesn't acknowledge in 5 minutes, escalate to the team lead. Still nothing after 15? Call the CTO. Every incident gets handled, even if someone's phone is on silent. Configurable timeouts, different channels at each step. Acknowledge to stop escalation, or let it keep climbing.
Build weekly, bi-weekly, or custom rotation schedules. Everyone takes their fair share. The schedule is always visible so your team knows exactly who's on call and when their turn comes around. Multiple schedules for different teams or services.
People take holidays. People get sick. People swap shifts because they have a dentist appointment on Tuesday. Overrides let you handle all of it without breaking the underlying rotation. One-time overrides for specific date ranges, and the original schedule stays intact.
Install one agent. Get security monitoring, infrastructure alerts, and on-call management - all included, every plan.