Operations¶

Use this page as the first day-2 runbook when you need to inventory a VM fleet, confirm current health, and decide which deeper operations procedure to execute next.

Prerequisites¶

Azure CLI installed and authenticated with az login.
Reader or higher access to the subscription that contains the VMs.
A target resource group already chosen for the current shift, exported as $RG.
At least one VM in scope so that $VM_NAME can point at a concrete machine for the quick health sample.

When to Use¶

At the start of an operations shift when you need a fast fleet baseline.
Before patching, resizing, backup work, or incident response so you know the current power state and recent control-plane changes.
When you inherit a VM estate and need to route work to the right detailed runbook in this section.

Procedure¶

Build the day-2 inventory snapshot¶

flowchart TD
    A[List in-scope VMs] --> B[Pick sample VM]
    B --> C[Check platform metrics]
    C --> D[Review recent control-plane changes]
    D --> E[Choose detailed runbook]
    E --> F[Execute change]
    F --> G[Re-verify health]

export RG="rg-vm-ops"
export VM_NAME="vm-app-01"
export VM_ID=$(az vm show --resource-group "$RG" --name "$VM_NAME" --query id --output tsv)

az vm list --resource-group "$RG" --show-details --query "[].{name:name,power:powerState,privateIp:privateIps,publicIp:publicIps,size:hardwareProfile.vmSize}" --output table

az monitor metrics list --resource "$VM_ID" --metric "Percentage CPU" --interval PT1H --aggregation Average --query "value[0].timeseries[0].data[-5:].{time:timeStamp,avg:average}" --output table

az monitor activity-log list --resource-group "$RG" --offset 1d --status Succeeded --query "[].{event:eventName.value,resource:resourceGroupName,caller:caller,time:eventTimestamp}" --output table

| Command | Purpose | | --- | --- | | az vm show | Resolves the VM resource ID used by later metric queries. | | --resource-group | Limits the lookup to the operations scope. | | --name | Selects the sample VM for metric inspection. | | --query | Returns only the resource ID. | | --output | Emits plain text that can be stored in $VM_ID. | | az vm list | Produces the initial fleet inventory for the resource group. | | --show-details | Adds power state and IP details to the VM inventory output. | | az monitor metrics list | Pulls recent platform metric data for the sample VM. | | --resource | Points the metric query at the VM resource ID. | | --metric | Requests CPU as the quick health signal. | | --interval | Narrows the chart to the last hour. | | --aggregation | Uses average CPU so you can spot sustained load. | | az monitor activity-log list | Shows recent successful control-plane operations in the same resource group. | | --offset | Restricts the activity log query to the last day. | | --status | Filters to completed operations. |

Expected result:

The VM inventory table shows the machines currently in scope and whether they are running, stopped, or deallocated.
The CPU query returns timestamped averages that tell you whether the selected VM is already under load.
The activity log identifies recent operator or automation actions before you touch the estate again.

Example output:

Name       PowerState   PrivateIp   PublicIp   Size
---------  -----------  ----------  ---------  ---------------
vm-app-01  VM running   10.20.1.4   52.x.x.x   Standard_D2s_v5

Route to the specific runbook¶

Use the command output to decide the next page:

New build or baseline drift -> Create and Configure VM
Access path or admin login issue -> Connect to VM
Capacity or storage pressure -> Manage Disks
Compute pressure or host fault suspicion -> Resize and Redeploy
Golden image, snapshot, or rollback preparation -> Snapshots and Images
Recovery point or restore request -> Backup and Restore
Monthly patch cycle or emergency patching -> Patching
Alert tuning or observability onboarding -> Monitoring and Alerting
Horizontal scaling or autoscale work -> VMSS Basics

Verification¶

Run one final check before moving on to a narrower runbook.

az vm list --resource-group "$RG" --show-details --query "[].{name:name,power:powerState,provisioning:provisioningState}" --output table

| Command | Purpose | | --- | --- | | az vm list | Reconfirms the VM fleet status before you begin a change. | | --resource-group | Uses the same operations scope as the inventory step. | | --show-details | Includes live power state information. | | --query | Returns only the fields needed for a go or no-go decision. | | --output | Formats the verification output as a readable table. |

Healthy verification looks like the expected VMs appearing with valid provisioning states such as Succeeded; if a VM is unexpectedly stopped or failed, switch to the detailed runbook that matches the problem before making unrelated changes.

Rollback / Troubleshooting¶

If az vm list returns nothing, confirm the correct subscription with az account show --output table and reset scope with az account set --subscription <subscription-id>.
If the metrics command returns no samples, the VM may be deallocated or you may be querying the wrong resource ID. Re-run az vm show and confirm the VM is running.
If the activity log is noisy, narrow the query to one VM by using --resource-id "$VM_ID" instead of the whole resource group.
If this page reveals a fault but not the fix, stop here and move to the specific runbook listed above rather than continuing with unrelated maintenance.