Outlook Activity Authentication Reset After Selective Pipeline Deployment #820
Description
We are encountering an issue in our PROD workspace related to pipeline deployments using the Publish All option in combination with the itemstobedeployed parameter.
Our setup includes QA, DBG, and PROD workspaces. When deploying a specific (dedicated) data pipeline using itemstobedeployed, the behavior is correct and consistent in QA and DBG environments — only the intended pipeline is deployed, and no other pipelines are impacted.
However, in the PROD workspace, although only the targeted pipeline is deployed, we are observing unexpected side effects on other existing pipelines. Specifically, pipelines that contain Outlook activities are getting their authentication reset. As a result, we are required to manually open each affected pipeline and reauthenticate the Outlook activity before the pipeline can run successfully.
An additional unusual behavior is that not all pipelines are impacted. The majority of pipelines with Outlook activities are getting their authentication reset, but some remain unaffected and retain valid authentication. We have confirmed that only the dedicated pipeline was included in the deployment, yet other pipelines are being partially modified (only the Outlook activity connection).
We are using itemstobedeployed specifically to avoid reauthenticating over 100+ pipelines containing Outlook activities during selective deployments. This behavior defeats that purpose and creates significant operational overhead.
For context, we are using two separate credentials: one dedicated for non-production environments (QA and DBG) and another dedicated for the PROD environment. While we believe this separation should not affect deployment behavior, we are mentioning it in case it is relevant for troubleshooting.
We would like to understand:
1> Why this issue is occurring only in the PROD workspace
2> Why only certain pipelines are affected while others remain intact
3> Whether this is a known issue related to selective deployments or connection handling in PROD
4> If there are recommended configuration changes or best practices to prevent authentication resets
Please advise on the root cause and any mitigation steps.