Spitfire
Self-hosted, no-code load and performance testing that finds your breaking point
Spitfire is a self-hosted, on-premise load testing and performance testing platform that shows you where your system breaks. Tests are set up in a form, no code required, and the load climbs step by step until a threshold breaks, showing the breaking point and the slowest steps with numbers.
It works in two modes. On watch, your critical flows run on production continuously at a low load (every 1 to 60 minutes with 1 to 10 virtual users) and notify you once when something changes, with SLO tracking, error budgets and a 30-minute close watch after each deploy. When needed, the same test runs as a full distributed load test before a sale, release or infrastructure change.
Tests go beyond HTTP, covering gRPC, WebSocket, Kafka, MQTT and SQL in the same test, and you can import OpenAPI, Postman, HAR, k6 or JMeter definitions. Version comparison runs the same test against two environments at once, corrects for environment differences, and can act as a CI release gate that fails slower releases. Every run becomes a shareable PDF, HTML report or time-limited link.
Because it runs entirely on your infrastructure, test data never leaves your network, the license works air-gapped, and runners dial out over TLS with key pinning so no inbound ports are needed. It includes a production write guard, SSO via OIDC and SAML, roles, an append-only audit log, AES-256-GCM encrypted secrets, and integrations with monitoring tools like Zabbix, Datadog, New Relic, Dynatrace and OpenTelemetry, plus Slack, Teams, Jira and CI systems. A free Android app lets you follow runs live and stop them with one tap.
It works in two modes. On watch, your critical flows run on production continuously at a low load (every 1 to 60 minutes with 1 to 10 virtual users) and notify you once when something changes, with SLO tracking, error budgets and a 30-minute close watch after each deploy. When needed, the same test runs as a full distributed load test before a sale, release or infrastructure change.
Tests go beyond HTTP, covering gRPC, WebSocket, Kafka, MQTT and SQL in the same test, and you can import OpenAPI, Postman, HAR, k6 or JMeter definitions. Version comparison runs the same test against two environments at once, corrects for environment differences, and can act as a CI release gate that fails slower releases. Every run becomes a shareable PDF, HTML report or time-limited link.
Because it runs entirely on your infrastructure, test data never leaves your network, the license works air-gapped, and runners dial out over TLS with key pinning so no inbound ports are needed. It includes a production write guard, SSO via OIDC and SAML, roles, an append-only audit log, AES-256-GCM encrypted secrets, and integrations with monitoring tools like Zabbix, Datadog, New Relic, Dynatrace and OpenTelemetry, plus Slack, Teams, Jira and CI systems. A free Android app lets you follow runs live and stop them with one tap.
