A cloud-based identity and access management service for securing user authentication and resource access
How to refresh user in MSAL cache
I recently updated my Active Directory password. Since then, I have been unable to run my Azure Machine Learning pipelines from VS Code. When running my code in the Azure CLI (pipeline using Azure ML SDK version: 1.53.0), an interactive authentication session is being launched, and I authenticate (successfully) through the pop up window in the browser. After doing so, the pipeline crashes with the message user ******@xxxx.com does not exist in MSAL token cache, run 'az login' .
This is despite me running an az login beforehand, or running it afterwards and try again. What has been attempted so far:
- Running a normal
az loginprior to running my pipeline - Running
az login --use-device-codebefore running pipeline - Running an
az account clearbefore a login - Running
az loginwhile explicitly providing the desired subscription - Setting the desired subscription to the default subscription before logging in.
- Logging in and out of Azure through the Azure portal
- Logging in and out of my MS account through edge.
- Restarting all programs (and PC) multiple times
- Signing in to Windows with new password, rather than PIN
- Manually removing the
msal.cachefile and repeating the above.
How do I refresh the MSAL token cache, so my Azure ML pipelines recognises the authentication? I have not had this issue before.