Cabot

Cabot is a self-hosted infrastructure monitoring and alerting service designed to deliver lightweight on-call paging without the overhead of enterprise incident platforms. Its central architecture couples service health checks directly with Graphite metrics and external calendar feeds, polling infrastructure status on premises rather than relying on external agent networks.

Why Cabot instead of Opsgenie

Opsgenie operates within Atlassian's cloud platform, providing hosted integrations, bidirectional Jira issue tracking, and cloud provider webhooks managed via mobile devices. Cabot takes an on-premises approach: it runs as an internal Python and Django service querying local Graphite metrics, Jenkins build jobs, and HTTP endpoints directly. Rather than exporting operational telemetry to a SaaS vendor, Cabot evaluates thresholds within private subnets, keeping monitoring data strictly behind local firewalls.

The operational divide centers on integration breadth and alerting depth. Opsgenie provides turnkey carrier routing, push notifications through mobile applications, and multi-tier escalation policies that hand off unanswered pages to secondary responders. Cabot instead delegates duty scheduling to an external iCal or Google Calendar feed, syncing shifts every thirty minutes and matching calendar event summaries to Django user accounts. It dispatches alerts through Twilio phone calls, SMS, email, or chat plugins like Slack, but omits complex escalation ladders and incident retrospectives. Teams gain a free alerting dispatcher connected to existing Graphite and Jenkins pipelines, trading away the managed integrations and mobile apps of commercial platforms.

What it does well

• Direct Graphite metric checks: Evaluates numeric thresholds against Graphite timeseries endpoints without running additional monitoring daemons on target hosts.
• Multi-channel notification routing: Dispatches alerts across Twilio voice calls, SMS messages, email, and team chat webhooks based on service severity levels.
• Calendar-based rota synchronization: Pulls on-call shifts from any iCal feed, enabling engineers to manage rotations inside Google Calendar.
• Jenkins pipeline verification: Continuously polls Jenkins build jobs and immediately triggers alerts when deployment runs fail.
• Web endpoint synthetic probing: Executes scheduled HTTP requests against internal microservices to validate response status codes and expected string content.