Overview of StackStorm: An Open‑Source Automation Platform

About

StackStorm is an open‑source platform for integrating and automating across services and tools. It links existing infrastructure and application environments, making it easier to automate them, with a particular focus on actions taken after events occur.

StackStorm helps automate common operational patterns. Examples include:

Convenient troubleshooting – triggering sensors from Nagios, Sentry, New Relic and other monitoring systems to capture failures, performing diagnostic checks on physical nodes, OpenStack or Amazon instances and application components, and publishing results to shared communication contexts such as HipChat or JIRA.

Auto‑remediation – detecting and verifying hardware failures on OpenStack compute nodes, evacuating instances appropriately, emailing administrators about potential downtime, and freezing workflows or invoking PagerDuty to wake human operators when problems arise.

Continuous deployment – using Jenkins to build and test, provisioning a new AWS cluster, opening some traffic with a load balancer, and performing forward or rollback based on New Relic application performance data.

StackStorm lets you combine these and other patterns into rules, workflows, or actions. These rules and workflows are stored as code, allowing the same collaborative development methods used for software projects, and they can be shared with the broader open‑source community via the StackStorm Exchange.

How It Works

Sensors are Python plugins that receive or monitor inbound or outbound events. When an event occurs in an external system and is processed by a sensor, a StackStorm trigger is emitted.

Triggers are StackStorm’s representation of external events. There are generic triggers (e.g., timers, webhooks) and integration triggers (e.g., Sentry alerts, JIRA issue updates). New trigger types can be defined by writing sensor plugins.

Actions are StackStorm’s outbound integrations. They include generic actions (ssh, REST calls), integration actions (OpenStack, Docker, Puppet) or custom actions. Actions can be Python plugins or any script, and can be used directly via CLI/API or as part of rules and workflows.

Rules map triggers to actions (or workflows), applying matching criteria and mapping trigger payloads to action inputs.

Workflows stitch actions together into “super‑actions”, defining order, conditional transformations, and data passing. Most automation tasks involve multiple steps, so workflows are essential.

Packs are units of content deployment. They group integrations (triggers and actions) and automation (rules and workflows), simplifying management and sharing of StackStorm extensible content. Many packs are available on the StackStorm Exchange, and users can create and share their own on GitHub.

Audit trails record the full context and results of action executions, whether manual or automatic, and are stored in audit logs for integration with external logging and analysis tools such as Logstash, Splunk, statsd, or syslog.

StackStorm has a modular architecture composed of loosely coupled services that communicate via a message bus and scale horizontally to deliver automation at scale. It provides a web UI, a CLI client, a full REST API, and Python client bindings to simplify developer work.

StackStorm is a new product under active development, and the community is encouraged to participate, provide feedback, and help shape its direction.

Original documentation: https://docs.stackstorm.com/overview.html