Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
In this tutorial, you use Azure CLI preview in a non-interactive pipeline job that signs in without a browser, deploys a template, and cleans up after itself.
When you finish, the job has signed in with a federated credential, deployed the template twice without creating duplicate resources, detected a failed command through its exit code, and removed both the test resources and its sign-in state.
In this tutorial, you:
- Sign in without a browser
- Prepare the job
- Deploy twice
- Detect a failure
- Clean up
Note
The preview supports managed identity and service principal authentication, but might not cover every automation scenario. Validate this walkthrough with your pipeline provider and preview build before using it. For more information, see Current limitations.
Prerequisites
- Bash or zsh on Linux or macOS, or PowerShell 7.6 or later on Windows.
- A test pipeline agent with the Azure CLI preview, installed as described in Set up and switch sign-in profiles.
- A test service principal with least-privilege access to the test subscription and a federated credential for your pipeline provider. The provider supplies a token file whose path is in
AZURE_FEDERATED_TOKEN_FILE. - The sample storage template from Deploy and troubleshoot a Bicep template.
Never print tokens, secrets, or the process environment in job output.
You need an Azure subscription. If you don't have one, create a free account.
Set variables
set -euo pipefail
tenant="<test-tenant-id>"
sub="<test-subscription-id>"
client_id="<application-client-ID>"
location="<region>"
template="<path-to-sample-storage.bicep>"
run=$(date +%Y%m%d%H%M%S)
profile="pipeline-$run"
rg="azclips-ci-$run"
deployment="ci-$run"
storage="clips$(openssl rand -hex 9)"
Sign in without a browser
federated_token="$(cat "$AZURE_FEDERATED_TOKEN_FILE")"
az login --service-principal --client-id "$client_id" --tenant-id "$tenant" --subscription-id "$sub" --federated-token "$federated_token" --out-profile "$profile"
The sign-in completes without a browser prompt and saves a profile for the job. The --out-profile parameter keeps the job's sign-in separate from other profiles on the agent.
Prepare the job
az profile show "$profile"
az group create --subscription-id "$sub" --name "$rg" --location "$location"
az profile show returns the job's profile, with the tenant and subscription you specified and the service principal authentication method. az group create returns the resource group with "provisioningState": "Succeeded".
Deploy twice
Pipelines often run the same job more than once, for example when a run is retried. Deploy the template twice, as a rerun would, and confirm that the second deployment reuses the same storage account.
ids=()
for attempt in 1 2; do
ids+=("$(az deployment group create --subscription-id "$sub" --resource-group "$rg" --name "$deployment" --template-file "$template" --parameter "storageAccountName=$storage" --mode Incremental --query properties.outputs.storageAccountId.value --output tsv)")
done
test "${ids[0]}" = "${ids[1]}"
echo "Both deployments returned the same storage account."
Both deployments succeed and return the same storage account resource ID, so the rerun doesn't create a second storage account. If the IDs differ, the job fails.
Detect a failure
if az group show --subscription-id "$sub" --name "$rg-missing" --output json; then
echo "Missing-resource lookup unexpectedly succeeded." >&2
exit 1
fi
The lookup fails with a not-found error and a nonzero exit code, and the job continues. A pipeline detects failed commands the same way: by checking the exit code. If the lookup unexpectedly succeeds, the job fails.
Clean up resources
Configure the pipeline to run cleanup even when an earlier step fails. The examples below show the cleanup commands, not a complete pipeline job or failure-handling wrapper.
az group delete --subscription-id "$sub" --name "$rg"
az group exists --subscription-id "$sub" --name "$rg"
az profile set --name "$profile"
az logout
az profile remove --name "$profile"
unset federated_token
az group exists returns false, and the job's profile is removed, so no sign-in state remains on the agent.
Other sign-in methods
Replace the sign-in example with one of the following. Keep secrets and certificates in the pipeline's secret store, never in source or job output.
Managed identity, on an agent with an assigned managed identity. Set the client ID to the managed identity's client ID.
- Azure CLI preview in Bash:
az login --identity --client-id "$client_id" --subscription-id "$sub" --out-profile "$profile" - Azure CLI preview in PowerShell:
az login --identity --client-id $clientId --subscription-id $sub --out-profile $profileName
Client secret
For client-secret sign-in, see Authentication. Store the secret in your pipeline's secret store. Service principal certificate sign-in isn't supported in this preview.
Related content
- For more information, see Get started with the preview.
- For more information, see Set up and switch sign-in profiles.
- For more information, see Recover from command failures with AI troubleshooting.