Unit 5 capstone — The week after an incident
AZ-104 › Unit 5 › Capstone
Unit 5 capstone — The week after an incident
Northwind Operations reviews a bad week
- AZ104-U5
- 2
- 13
- 5
- 25 minutes
The brief
Northwind's operations team had a bad week. A line-of-business application degraded last Tuesday and nobody can say why: the resource involved has never had a diagnostic setting. A CPU alert covering thirty virtual machines produced one ticket when the team expected thirty, and a second alert rule has been firing hourly for a month — the on-call engineer wants it silenced tonight without losing the rule. An analyst deleted a file from a file server two days ago. And the business has asked what would happen if the region hosting the application became unavailable.
What they need
- Investigate last Tuesday's degradation on a resource that has never had a diagnostic setting.
- Get one alert per virtual machine from a CPU rule covering thirty of them.
- Silence a noisy alert rule tonight without deleting or disabling it.
- Recover a file deleted from a file server two days ago.
- Keep the application available if its region becomes unavailable.
Commit before you read on
Write down your answer to each of the 5 numbered requirements above — the mechanism, and the scope you would apply it at. Then open the working below and compare. Reading the reasoning beside the question is the one way to feel like you knew it without having known it.
▶Show the working
1. Investigate last Tuesday with no diagnostic setting
Use the platform metrics and the activity log. The resource logs for last Tuesday do not exist and cannot be produced.
This is the honest answer and the one the exam wants. Platform metrics and the activity log are automatically collected without configuration — a diagnostic setting only sends them somewhere else. Resource logs need the setting to be collected at all, and collection starts when the setting does. Creating one now fixes next Tuesday. Any option that recovers last week's resource logs is describing something impossible.
2. One alert per machine across thirty
Nothing to change — that is already the behaviour. Investigate why it did not happen.
If you're monitoring more than one resource, the alert rule condition is evaluated separately for each resource, and alerts are fired for each resource separately. So a single ticket for thirty machines points at the scope of the rule or at the action group's grouping, not at a missing per-resource setting. Reading the requirement as "add a rule per machine" is the expensive wrong answer.
3. Silence a rule tonight without losing it
An alert processing rule that suppresses notifications for that window.
Three objects, three jobs: the alert rule detects, the action group notifies, and the alert processing rule suppresses or redirects. Disabling the rule loses the detection; deleting the action group affects every rule using it. And note what cannot be done: fired alert instances are read-only and cannot be edited — configuration changes apply only to future alerts, so the hour's existing noise stays.
4. Recover a file deleted two days ago
Restore from Azure Backup.
Backups are stored in a Recovery Services vault with built-in management of recovery points, and a restore is how a deleted item comes back. Site Recovery is the wrong service here however plausible it looks in the option list — replicating a deletion is all it would have done. Budget for the restore: outbound data charges apply to data transferred from a vault during a restore.
5. Survive the region becoming unavailable
Azure Site Recovery.
Site Recovery keeps business apps and workloads running during outages by replicating to a secondary location; when an outage occurs you fail over and access apps from there, and fail back once the primary is running again. It answers availability, not recovery — which is why requirements 4 and 5 need two different services and neither substitutes for the other.
The reasoning this unit rewards
Check yourself
Where these figures come from
Every figure above was read from the raw documentation below on the day this sheet was written. The sha1 is git hash-object over the bytes as fetched, so a doc that changes underneath this sheet can be detected rather than assumed.
| Document | sha1 |
|---|---|
| Diagnostic settings in Azure Monitor | 123baaa50c7f |
| What are Azure Monitor alerts? | 7080cb554190 |
| What is Azure Backup? | 57bc50276ab0 |
| About Site Recovery | 1002240a77cf |