Amazon Web Services → Microsoft Azure
Azure equivalent of Amazon S3
The Azure equivalent of Amazon S3 is Azure Blob Storage.
About Amazon S3
| Provider | Amazon Web Services |
|---|---|
| Status | Generally available |
| What it is | Object storage in buckets, with 11 nines of durability, a choice of storage classes, versioning, replication, Object Lock, and fine-grained IAM and bucket policies. General purpose buckets hold any object type; directory, table, and vector buckets are specialized bucket types. |
| Documentation | docs.aws.amazon.com |
| Pricing | Official pricing page |
Azure mappings (2)
| Capability | Azure equivalent | Match |
|---|---|---|
| Object storage | Azure Blob Storage | Direct match |
| Data lake storage with real directories | Azure Data Lake Storage (hierarchical namespace) | Partial match |
Object storage(1)
Object storage Direct match
The Azure equivalent of Amazon S3 is Azure Blob Storage. Both store objects in flat containers (S3 buckets, Blob containers) with an HTTP API, multiple cost tiers, versioning, replication, and WORM locking.
- Account model. S3 buckets are top-level resources with globally unique names. Blob containers live inside a storage account, and the account sets the redundancy, default access tier, network rules, and some features (such as the hierarchical namespace) for every container in it.
- Redundancy. S3 Standard stores data across at least three AZs by default; One Zone classes and Express One Zone use a single AZ. On Azure you choose per account. LRS is a single datacenter, ZRS spans zones, and GRS/GZRS add an asynchronous copy in the paired region. Pick ZRS explicitly if you want S3-like zone resilience.
- Object types. S3 has one object type (with multipart upload). Blob Storage has block blobs (the S3 equivalent), append blobs (for logs), and page blobs (random-access, used by legacy unmanaged disks).
Object access, file access & analytics(1)
Data lake storage with real directories Partial match
On AWS, data lakes are built directly on S3 general purpose buckets. On Azure, you turn on the hierarchical namespace for a storage account to get Azure Data Lake Storage, which adds real directories, atomic renames, and POSIX-style ACLs.
- Namespace. S3 general purpose buckets are flat, and folders are key prefixes, so "renaming" a folder copies every object. ADLS directories are real, with atomic rename and delete. (S3 directory buckets do have a hierarchical namespace, but only for Express One Zone.)
- Permissions. ADLS supports POSIX-like ACLs on directories and files as well as RBAC. On AWS, fine-grained data lake permissions usually come from Lake Formation or S3 Access Grants.
- Feature trade-offs. The hierarchical namespace is required for SFTP and NFS 3.0 on Blob, but some Blob features have limits on ADLS accounts. Check the feature support matrix before enabling it.