The environment of Nordlicht Reisen GmbH, a fictional travel agency, is in place (lesson 6.1). From now on the work begins that is not in any template: checking that everything runs, classifying anomalies, and clearing away legacy leftovers. This lesson sums up the course from the operations point of view. It brings no new exam material but organizes the known material: an operations round, a table of typical error patterns, and an outlook on where you can go from here.
Prior knowledge: Everything in this lesson points to the lessons of the course; if you no longer remember a term, look it up there.
The operations round
Microsoft prescribes no fixed round. What follows is a suggestion that you adapt to your environment. Each row names what you check, with what, and where it is in the course.
Daily
| What | With what | Lesson |
|---|---|---|
| Look at new and open alerts, acknowledge or close them | Portal: Monitor → Alerts | 5.3 |
| Backup jobs of last night | az backup job list --resource-group <gruppe> --vault-name <tresor> --output table |
5.5 |
| Messages from Azure about outages and maintenance | Service Health | 5.3 |
Weekly
| What | With what | Lesson |
|---|---|---|
| Noncompliant resources | az policy state list --resource-group <gruppe> --filter "(isCompliant eq false)" |
1.5 |
| Costs against the budget, recommendations from Azure Advisor | Portal: Cost Management, Advisor | 1.6 |
| Alerts that nobody ever handles, and rules that fire too often | Alert list, adjust rules | 5.3 |
| Storage accounts: availability, capacity | Storage insights | 5.4 |
Monthly
| What | With what | Lesson |
|---|---|---|
| Review role assignments: who has which role on which scope | az role assignment list --all --output table |
1.3 |
| Test a restore: a disk from a recovery point | Portal or az backup restore restore-disks |
5.5 |
| Run a test failover and clean it up, where Site Recovery is running | Portal | 5.6 |
| Expiring keys, SAS tokens, and certificates | Storage account, App Service | 2.2, 3.8 |
| Quotas and available sizes | az vm list-skus, quota display |
0.1 |
| Orphaned resources: disks without a machine, groups without a purpose | az resource list --resource-group <gruppe> --output table |
1.4 |
Why the restore and the test failover should be regular: a backup that was never restored is an assumption. The rehearsal shows whether permissions, networks, and names are right when it counts.