What's changed: Deepened AZ-104 Chapter 2 to Associate depth (tables, scenarios, FAQ, traps; localized figures)
2.1Storage Accounts and Redundancy
Understand the four services a storage account groups (Blob, Files, Queue, Table) and the redundancy options that protect data (LRS / ZRS / GRS).
The entry point for storing data on Azure is the storage account. One account groups four services—Blob, Files, Queue, Table—chosen by use case.
2.1.1Storage account services
- Blob: object storage for unstructured data (images, logs, backups).
- Files: managed file shares mountable over SMB/NFS by VMs and on-prem.
- Queue: passes messages between components. Table: key-value NoSQL.
2.1.2Redundancy options
- LRS (locally redundant): 3 copies within one datacenter; cheapest but vulnerable to a DC failure.
- ZRS (zone redundant): spread across availability zones in the same region; resilient to a zone failure.
- GRS (geo-redundant): replicates to another region; protects against a region-wide disaster.
A common AZ-104 distinction: LRS = within one DC, ZRS = multiple zones in the same region, GRS = replicate to another region. If region-disaster protection is required, choose GRS.
A storage account groups four services (Blob/Files/Queue/Table); choose a performance tier (Standard = HDD/general, Premium = SSD/low-latency) and redundancy at creation. Think of redundancy on two axes—"in-region redundancy" and "cross-region replication." LRS = 3 copies within one datacenter (cheapest, weak to a DC failure); ZRS = spread across three availability zones in the region (resilient to a zone failure). Adding geo replication gives GRS = async copy to another region (LRS at two sites) and GZRS = ZRS primary plus cross-region copy. The geo secondary is normally not readable; to read it choose RA-GRS/RA-GZRS (read-access geo-redundant). Switch over on a regional outage via account failover (async replication, RPO ~minutes). The standard account kind is General Purpose v2. The axes: "cheapest/single-DC = LRS," "zone-failure protection = ZRS," "region-disaster protection = GRS/GZRS," "read the secondary = the RA- variant."
| Redundancy | Protects against |
|---|---|
| LRS | 3 copies in one DC (in-DC failure) |
| ZRS | AZ failure in the same region |
| GRS / GZRS | Region-wide disaster (cross-region copy) |
| RA-GRS / RA-GZRS | Above + readable secondary |
Scenario: tolerate a zone failure and also read from the secondary during a region disaster. Use ZRS for in-region zone redundancy, GZRS to add geo replication for region-disaster protection, and RA-GZRS if you must read the secondary. If cost is paramount and a single DC suffices, use LRS. Switch regions via account failover.
Q. Cheapest? LRS. Q. Zone-failure protection? ZRS. Q. Region-disaster protection? GRS/GZRS. Q. Read the secondary? RA-GRS/RA-GZRS. Q. SSD/low-latency? Premium. Q. Standard account kind? General Purpose v2.
Watch the mix-ups: (1) ZRS (zones in one region) and GRS (another region) protect different things—ZRS for zone failures, GRS for region disasters. (2) The geo secondary isn’t readable by default—use the RA- variant. (3) Geo replication is async, so RPO isn’t zero. (4) Some redundancy changes are restricted after account creation (especially Premium).
Choose by purpose: Blob is object storage for unstructured data, while Files is file storage you mount and share over SMB/NFS.
2.1.3Section summary
- One account = four services: Blob / Files / Queue / Table
- Redundancy: LRS (in-DC) → ZRS (zones) → GRS (another region)
Sign in to track progress — Log in.
Quick check
(just a quick review)Q1. Which Azure Storage service best stores unstructured data like images and logs?
Q2. You want data replicated to another region to survive a region-wide disaster. Which redundancy?
Q3. Which redundancy spreads data across availability zones within the same region?