S
Satisfied Lark
Description:
When executing IACM pipelines across multiple workspaces (e.g., via matrix, looping, or multi-workspace execution steps), a pipeline failure in one workspace currently requires re-running the entire pipeline for all targeted workspaces.
We request the ability to selectively retry/re-run only the failed workspace steps or failed iterations within a multi-workspace pipeline run, while skipping or preserving the successful execution state of the workspaces that passed.
Business Impact:
  1. Cost & Time Efficiency: Significantly reduces total pipeline execution duration and resource consumption by preventing unnecessary, repetitive executions on workspaces that have already succeeded.
  2. Developer Productivity: Enables engineers to quickly fix an issue in a specific failing workspace and re-trigger execution immediately without waiting for long, unchanged multi-workspace plans/applies to complete.
  3. Blast Radius Reduction: Decreases unnecessary API calls and state lock operations against cloud providers and external systems by avoiding redundant runs on stable infrastructure.