What happens if the Knapsack Pro API is down?

Ruby (Regular Mode)

Knapsack Pro will retry to connect with the API multiple times before switching to Fallback Mode, printing a warning, and splitting test files across CI nodes based on test file names.

In the unlikely scenario where some CI nodes start in Fallback Mode but others don't, some tests might be skipped. If this is not acceptable, you can disable Fallback Mode with KNAPSACK_PRO_FALLBACK_MODE_ENABLED=false.

If your CI provider allows retrying only one parallel CI node then please read about KNAPSACK_PRO_FIXED_TEST_SUITE_SPLIT.

Ruby (Queue Mode)

Knapsack Pro will retry to connect with the API multiple times before switching to Fallback Mode, printing a warning, and splitting test files across CI nodes based on test file names.

In the unlikely scenario where some CI nodes switch to Fallback Mode but others don't, some tests might be run on multiple CI nodes. In other words, in Queue Mode you have the guarantee of at-least-once tests execution.

If your CI provider allows retrying only one parallel CI node then please read about Dynamic Split vs Fixed Split.

Cypress and Jest

Knapsack Pro will retry to connect with the API multiple times before switching to Fallback Mode, printing a warning, and splitting test files across CI nodes based on test file names.

In the unlikely scenario where some CI nodes switch to Fallback Mode but others don't, some tests might be run on multiple CI nodes. In other words, in Queue Mode you have the guarantee of at-least-once tests execution.

If your CI provider allows retrying only one parallel CI node then please read about KNAPSACK_PRO_FIXED_QUEUE_SPLIT (Cypress, Jest).