Updated resource deletion instructions

docs/deployment-instructions.md

@@ -102,16 +102,18 @@ For further learning, you can run through some of the demo scripts listed below:
 
 ## Cloud costs and cleanup
 
-Once you are done deploying, testing, exploring, you should delete the provisioned resources to prevent incurring additional costs.
+Once you are done deploying, testing, exploring, you should delete the `contoso-traders-rg{SUFFIX}` resource group to prevent incurring additional costs.
 
-Once done, you can safely delete the `contoso-traders-rg{SUFFIX}` resource group. The `contoso-traders-aks-nodes-rg{SUFFIX}` will be automatically deleted as part of the AKS cluster deletion.
+  ![resource group deletion](./images/resource-group-deletion.png)
+
+The `contoso-traders-aks-nodes-rg{SUFFIX}` will be automatically deleted as part of the AKS cluster deletion.
 
 >A quick note on costs considerations when you deploy the application to your Azure subscription:
 >
 > 1. Azure Load Testing ([pricing details](https://azure.microsoft.com/pricing/details/load-testing/)): The number of virtual users and duration of the test are the key factors that determine the cost of the test. In this demo, the load tests are configured to use 5 virtual users and the test is set to run for 3 mins.
-> 2. Azure Kubernetes Service ([pricing details](https://azure.microsoft.com/pricing/details/kubernetes-service/)): The number of nodes and the number of hours that the cluster is running are the key factors that determine the cost of the cluster. In this demo, the cluster is configured to use 1 node (powered by vm scale sets) and the cluster is set to run 24x7. You can manually stop the cluster when not in use.
+> 2. Azure Kubernetes Service ([pricing details](https://azure.microsoft.com/pricing/details/kubernetes-service/)): The number of nodes and the number of hours that the cluster is running are the key factors that determine the cost of the cluster. In this demo, the cluster is configured to use 1 node (powered by vm scale sets) and the cluster is set to run 24x7 (you can manually stop the cluster when not in use). Because of a [limitation in the AKS bicep schema](https://github.com/Azure/bicep/issues/6974), the AKS cluster has to use premium SSD storage disks.
 > 3. Azure Container Apps ([pricing details](https://azure.microsoft.com/pricing/details/container-apps/)): Each instance has 0.5 vCPU and 1.0 GiB of memory. In this demo, the container app is configured to use 1 instance, but can autoscale out to max 3 instances under load.
-> 4. Azure Virtual Machines ([pricing details](https://azure.microsoft.com/pricing/details/virtual-machines/windows/)): The jumpbox VM uses the `Standard_D2s_v3` VM size, which has 2 vCPU and 8 GiB of memory. The jumpbox VMs are set to run 24x7. You can manually stop the VM when not in use.
+> 4. Azure Virtual Machines ([pricing details](https://azure.microsoft.com/pricing/details/virtual-machines/windows/)): The jumpbox VM uses the `Standard_D2s_v3` VM size, which has 2 vCPU and 8 GiB of memory. The jumpbox VMs are schedule to auto-shutdown at 1900 UTC daily. You can also manually stop & deallocate the VM when not in use.
 > 5. Github Actions / storage quota ([pricing details](https://docs.github.com/en/billing/managing-billing-for-github-actions/about-billing-for-github-actions#included-storage-and-minutes)): We've set the playwright test to enable recordings only on failures/retries. This brings the playwright report to ~55 MB when tests fail.
 >
 > The above costs are based on the default configuration of the demo. You can modify the configuration to reduce the costs. For example, you can reduce the number of instances in the container app, reduce the number of virtual users in the load test, etc.