Artillery
Load testing and smoke testing toolkit for modern backends
Artillery adds a single-request diagnostic command and finally makes ESM a first-class citizen.
◆Recent moves
- 4d ago
artillery ping lands, plus full ESM and complete TypeScript definitions
⚡ SPARKThe first release in this window to add capability rather than repair it: a ping command for single-request diagnostics, ESM support throughout the extension points, and TypeScript types covering the whole script format. It also continues the worker-image consolidation arc by bundling the fake-data plugin instead of downloading it at test startup.
View source ↗ - 2mo ago
Playwright 1.61 and a Faker swap in fake-data
A Playwright bump and a replacement of Falso with Faker.js inside the fake-data plugin. Maintenance in the cadence that dominates this window.
View source ↗ - 3mo ago
http.timeout no longer silently capped at 8 seconds
Fixes a cap that made requests to slow origins fail with ERR_SOCKET_TIMEOUT regardless of the configured http.timeout, plus startup MODULE_NOT_FOUND errors and Fargate runs starting fewer VUs than configured. The timeout bug is the consequential one — it produced failures that looked like the system under test was broken.
View source ↗ - 3mo ago
W3C trace context propagation and a Got 11 to 14 engine upgrade
Adds traceparent headers on outgoing requests so backend traces correlate with the test that caused them, and moves the HTTP engine from Got 11 to Got 14 — flagged as potentially breaking for tests that reach into Got internals. Fargate workers also self-terminate if the CLI disappears for 180 seconds, closing a runaway-cost gap.
View source ↗ - 6mo ago
Playwright 1.58.1 and a Fargate binary mismatch fix
Bumps Playwright and ensures Fargate uses pre-packaged Playwright and CLI versions rather than reinstalling them. Small, but part of the worker-image consolidation that runs through this window.
View source ↗ - 6mo ago
Playwright 1.58 and a Fargate metrics completeness fix
A Playwright bump plus a fix for incomplete results when a test tracks many distinct metrics on Fargate. Routine upkeep of the distributed runner.
View source ↗