What's changed: Deepened AZ-305 Chapter 3 (added comparison tables, scenarios, FAQs, exam traps, deep-dive paragraphs to each section; localized figures to Japanese)
3.1Designing Backup and Data Protection
Understand data protection design—Azure Backup, recovery points (RPO)/retention, soft delete/immutable storage, and geo-redundant storage (GRS). Recover reliably from data loss or mistakes.
Business continuity starts with data protection. Centralize backups with Azure Backup and prepare for regional failures with storage redundancy.
3.1.1Backup and redundancy
- Azure Backup: policy-based scheduled backups of VMs/SQL/Files; retains recovery points.
- Storage redundancy: LRS (in-DC) < ZRS (across zones) < GRS/RA-GRS (another region); use GRS for regional failures.
- Soft delete/immutable: soft delete recovers accidental deletes; WORM prevents tampering.
- RPO/retention: design backup frequency/generations by tolerable data loss and retention.
Common on AZ-305: centralized backup = Azure Backup (Recovery Services vault), regional protection = GRS/RA-GRS, zone resilience = ZRS, recover accidental deletes = soft delete, tamper-proof = immutable storage (WORM). Choose RA-GRS when you also need read access to the geo-replica.
AZ-305 data-protection design probes "which service and configuration meet RPO, retention, redundancy, and tamper-resistance requirements." Azure Backup protects workloads via a Recovery Services vault (VMs, SQL in VM, Azure Files) and a Backup vault (newer targets like Azure SQL DB, Blob, disks), designing frequency, retention (daily/weekly/monthly/yearly GFS), and instant restore through a backup policy. The vault itself gains resilience to ransomware/malicious deletion via soft delete and immutable vault / MUA (multi-user authorization), and prepares for regional failures with a geo-redundant vault. Storage redundancy is chosen by failure-domain breadth: LRS (3 copies in one DC) < ZRS (across zones) < GRS (async-replicate to another region; not readable by default) < RA-GRS (geo replica readable). For data-level protection, combine Blob/Files soft delete, Blob versioning, and immutable storage (time-based/legal-hold WORM). The key is to work backward from business RPO/retention/compliance to size backup frequency, retention generations, redundancy level, and tamper resistance without over- or under-engineering.
| Redundancy | Replication scope | Survives |
|---|---|---|
| LRS | 3 copies in one DC | Disk/node failure |
| ZRS | Multiple zones in region | Zone (DC) failure |
| GRS | Async to another region | Regional failure (not readable) |
| RA-GRS | GRS + secondary readable | Regional failure + read continuity |
Scenario: Protect production VMs and file shares from ransomware, accidental deletion, and regional disaster—and ensure the backups themselves cannot be deleted. → Configure daily backups with GFS retention in Azure Backup, and enable soft delete and immutable vault / MUA (multi-user authorization) on the vault. Use a geo-redundant vault for regional failures. For Blob, prevent tampering/deletion with versioning + immutable storage (WORM).
FAQ: GRS vs RA-GRS? Both async-replicate to another region, but GRS does not normally allow reading the secondary (accessible on Microsoft-initiated/account failover), while RA-GRS keeps the secondary readable at all times, enabling read continuity during a regional failure or geographically closer reads. Choose RA-GRS when read continuity is required.
Exam trap: Leaving storage at LRS when you need to "survive a zone failure" is wrong—LRS is in-DC redundancy; a zone failure needs ZRS or higher. Also, assuming "we take backups" satisfies ransomware protection is wrong—you need soft delete + immutable vault (MUA) so the backups themselves cannot be deleted/encrypted.
3.1.2Section summary
- Protection = Azure Backup + storage redundancy (LRS/ZRS/GRS)
- Recovery = soft delete / immutable storage (WORM)
Sign in to track progress — Log in.
Quick check
(just a quick review)Q1. To prepare for a regional failure, you want storage auto-replicated to another region. Which redundancy?
Q2. Schedule backups of VMs/SQL under common policies and centrally manage recovery points. What?
Q3. You want to restore accidentally deleted Blobs within a window. What do you enable?