Last updated: May 2026

Azure Storage Intermediate AZ-104 ⏱ 13 min read

Azure Storage Redundancy — LRS, ZRS, GRS, GZRS

Azure Storage always keeps multiple copies of your data to protect against failures. But how many copies, where they're stored, and what failures they protect against varies significantly across the four redundancy options. Choosing the wrong one means either overpaying for protection you don't need, or under-protecting data that could be catastrophically lost.

What you'll learn LRS, ZRS, GRS, GZRS and RA variants explained in depth · What each option protects against · Durability figures · Cost comparison · Changing redundancy after creation · Geo-failover explained · Choosing the right option for each scenario

Table of Contents

LRS — Locally Redundant Storage
ZRS — Zone Redundant Storage
GRS — Geo Redundant Storage
GZRS — Geo-Zone Redundant Storage
RA-GRS and RA-GZRS
Full Comparison Table
Changing Redundancy
Geo-Failover
Choosing the Right Option
Practice Questions

LRS — Locally Redundant Storage

LRS replicates your data 3 times within a single data centre in your chosen region. All three copies are in the same physical building, on different racks.

Property	Value
Number of copies	3
Locations	Same data centre, different racks
Protects against	Hardware failures, drive failures, rack failures
Does NOT protect against	Data centre fire/flood, regional disaster
Durability	99.999999999% (11 nines)
Cost	Lowest

⚠️

LRS Risk If the entire data centre is destroyed (fire, flood, power loss), all 3 copies are lost. Microsoft recommends LRS only for data that can be reconstructed easily or for non-critical scenarios.

ZRS — Zone Redundant Storage

ZRS replicates your data synchronously across 3 Availability Zones in the same region. Each copy is in a physically separate data centre with independent power, cooling, and networking.

Property	Value
Number of copies	3
Locations	3 Availability Zones in same region
Protects against	Hardware failures + entire data centre (zone) failure
Does NOT protect against	Regional disaster (all zones affected)
Durability	99.9999999999% (12 nines)
Cost	~25% more than LRS

💡

ZRS is the recommended default for production For most production workloads in regions that support AZs (like Central India), ZRS is the best balance of protection and cost. It protects against data centre failures — the most common serious failure — without the cost and complexity of geo-redundancy.

GRS — Geo Redundant Storage

GRS replicates data to a paired region hundreds of kilometres away. Within each region, it uses LRS (3 copies). Total copies: 6.

Property	Value
Number of copies	6 (3 primary + 3 secondary)
Locations	Primary region (LRS) + paired secondary region (LRS)
Protects against	Hardware failures + regional disasters
Secondary readable?	Only after failover (not by default)
Durability	99.99999999999999% (16 nines)
Cost	~2x LRS

ℹ️

Replication is Asynchronous GRS replication to the secondary region is asynchronous — there may be a lag of minutes to hours. If the primary region fails before all data is replicated, some recent writes may be lost (RPO is not zero).

Azure Region Pairs

GRS always replicates to the designated paired region — you cannot choose the secondary region:

Primary Region	Paired Secondary
Central India (Pune)	South India (Chennai)
East US	West US
UK South	UK West
West Europe	North Europe

GZRS — Geo-Zone Redundant Storage

GZRS combines the best of ZRS and GRS — ZRS in the primary region, LRS in the paired secondary region. Maximum protection.

Property	Value
Number of copies	6 (3 across AZs primary + 3 LRS secondary)
Locations	3 AZs in primary + LRS in paired secondary
Protects against	Zone failures + regional disasters
Durability	99.99999999999999% (16 nines)
Cost	Highest

RA-GRS and RA-GZRS

The RA (Read-Access) variants allow reading from the secondary region at any time — even when the primary is healthy. This is useful for globally distributed read workloads.

Secondary endpoint: https://[accountname]-secondary.blob.core.windows.net
Secondary data may be slightly stale (asynchronous replication lag)
Cannot write to the secondary — read-only

Full Comparison Table

Option	Copies	Zone Failure	Regional Disaster	Secondary Readable	Cost
LRS	3	❌ No	❌ No	—	$
ZRS	3	✅ Yes	❌ No	—	$$
GRS	6	❌ No	✅ Yes	After failover only	$$$
GZRS	6	✅ Yes	✅ Yes	After failover only	$$$$
RA-GRS	6	❌ No	✅ Yes	✅ Always	$$$+
RA-GZRS	6	✅ Yes	✅ Yes	✅ Always	$$$$+

Changing Redundancy

You can change redundancy on an existing storage account without downtime. Some transitions are instant, others require a data migration:

Transition	Method
LRS → ZRS	Live migration (request via support) or manual
LRS/ZRS → GRS/GZRS	Instant — portal or CLI
GRS → LRS	Instant — but loses geo-redundancy

Azure CLI Change storage account redundancy

az storage account update \
  --name mystorageaccount2026 \
  --resource-group myResourceGroup \
  --sku Standard_GRS

Geo-Failover

If the primary region becomes unavailable with GRS/GZRS, you can initiate a failover — the secondary region becomes the primary. After failover:

Your storage account DNS points to the secondary (now primary) region
The account becomes LRS in the new region (only 3 copies until geo-replication is re-established)
Some recent data may be lost (async replication lag)
Microsoft can also initiate failover on your behalf in a major outage

Choosing the Right Option

Scenario	Recommended
Dev/test, easily reconstructable data	LRS
Production data in AZ-supported region	ZRS
Data requiring regional DR, no AZ requirement	GRS
Critical data needing both zone + regional DR	GZRS
Global read performance + regional DR	RA-GRS or RA-GZRS
Regulatory compliance requiring geo-redundancy	GRS or GZRS

💡

AZ-104 Exam Tip This is one of the most heavily tested storage topics. Know what each option protects against, how many copies each stores, that GRS replication is asynchronous (not zero RPO), that RA-GRS allows reading the secondary without failover, and the Azure region pairs.

📝 Practice Questions

Click an option to check your answer. AZ-104 style questions.

Q1. A company stores critical financial records and needs protection against both data centre (zone) failures and regional disasters. Which redundancy option should they choose?

A LRS

B ZRS

C GRS

D GZRS

Q2. What is the difference between GRS and RA-GRS?

A RA-GRS replicates to two secondary regions instead of one

B RA-GRS allows reading from the secondary region at any time; GRS only after failover

C RA-GRS replicates data faster to the secondary region

D RA-GRS stores more copies of data than GRS

Q3. If a storage account using GRS fails over to the secondary region, what redundancy level does it temporarily have?

A LRS — only 3 local copies until geo-replication is re-established

B GRS — it automatically establishes a new secondary immediately

C ZRS — Azure automatically distributes across zones

D No redundancy — single copy only

Q4. How many total copies of data does GRS store?

A 3

B 6

C 9

D 12

Q5. Which Azure region is the designated pair for Central India (Pune)?

A West India (Mumbai)

B South India (Chennai)

C Southeast Asia (Singapore)

D East Asia (Hong Kong)

← Previous Azure Table Storage Next → Shared Access Signatures

Comments

Disclaimer: RedKite Cloud is an independent educational resource and is not affiliated with, endorsed by, or officially connected to Microsoft Corporation. All product names, logos, and trademarks are property of their respective owners.