From e0fdac036094278f7358daf99ba9fef36026865a Mon Sep 17 00:00:00 2001
From: Pieter Noordhuis <pieter.noordhuis@databricks.com>
Date: Fri, 29 May 2026 14:04:12 +0200
Subject: [PATCH] Fix flaky bundle generate pipeline_and_deploy test: unique
 pipeline name

The pipeline name was set to "test-pipeline-${UNIQUE_NAME}" inside a
single-quoted JSON block, so bash never expanded ${UNIQUE_NAME} and every
run created a pipeline with the identical literal name. On the cloud variant
the Pipelines API rejects duplicate names, causing intermittent failures
when variants ran concurrently or a prior run's pipeline lingered.

Break out of the single quotes so $UNIQUE_NAME is interpolated, matching the
existing '${CURRENT_USER_NAME}' convention in the same script.

Co-authored-by: Isaac
---
 acceptance/bundle/generate/pipeline_and_deploy/script | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/acceptance/bundle/generate/pipeline_and_deploy/script b/acceptance/bundle/generate/pipeline_and_deploy/script
index db8e519574..cd04723111 100644
--- a/acceptance/bundle/generate/pipeline_and_deploy/script
+++ b/acceptance/bundle/generate/pipeline_and_deploy/script
@@ -4,7 +4,7 @@ trace $CLI workspace import "/Workspace/Users/${CURRENT_USER_NAME}/test.py" --fi
 
 title "Create a pipeline that references the files"
 PIPELINE_ID=$($CLI pipelines create --json '{
-  "name": "test-pipeline-${UNIQUE_NAME}",
+  "name": "test-pipeline-'${UNIQUE_NAME}'",
   "libraries": [
     {
       "notebook": {