Files
rustfs/rustfs
Martin Claesson c47c8f9f74 fix(sftp): allow bucket-scoped root listings (#8998)
* fix(sftp): allow bucket-scoped root listings

Listing the SFTP root required s3:ListAllMyBuckets. Without it a
bucket-scoped user got PermissionDenied on `ls /`; with it they saw every
bucket on the instance. S3 ListBuckets and the WebDAV root (#6298)
already fall back to a per-bucket filtered listing.

When the global ListAllMyBuckets check is denied, list buckets through
the session-aware ListBuckets path instead, so IAM filters the result to
buckets the principal may ListBucket or GetBucketLocation. Principals
with ListAllMyBuckets keep the unfiltered listing, and IAM outages and
timeouts still fail as before.

list_buckets_for_session and its request builder were WebDAV-only; they
are now also compiled with the sftp feature.

* fix(protocols): re-check the session key for bucket-scoped listings

list_buckets_for_session built its request from the identity captured at
login. Through the SFTP root-listing fallback, a service account disabled
or expired after login could keep listing bucket names, unfiltered when
its parent may list every bucket.

Validate the session key with check_key_valid_with_context, as
authorize_protocol_request does for every other protocol operation, build
the request from the fresh credentials and is_owner, map an invalid key
to AccessDenied, and attach the server context.

The SFTP fallback now passes the session's request headers and transport
security, like every other SFTP authorization. DummyBackend fails closed
when no session listing is queued, matching the trait default, and gains
a stall_authorization hook. New tests cover a service account disabled
mid-session, IAM outage and timeout on the global check (no fallback),
non-AccessDenied backend errors, truncation of fallback output and the
fail-closed default.

* test(protocols): compare error code without referencing s3s

The session-listing test compared against s3s::S3ErrorCode, which made
policy_test.rs a new s3s-referencing file and tripped the s3s footprint
ratchet (210 files against a baseline of 209). Compare the code's string
form instead.

* fix(protocols): refresh session bucket discovery without stale copies

Build bucket listing requests directly from freshly validated credentials
and server-scoped ownership instead of cloning and replacing login state.
Reuse the same authorization path for WebDAV capacity discovery.

Preserve typed access-denied checks through the protocol facade, and cover
request conditions, session-policy updates, revocation, listing bounds,
and fallback deadlines with regression tests.

Co-Authored-By: heihutu <heihutu@gmail.com>
Co-Authored-By: zhi22915 <qiuzgang@gmail.com>

---------

Co-authored-by: Hauser <housemecn@gmail.com>
Co-authored-by: heihutu <heihutu@gmail.com>
Co-authored-by: zhi22915 <qiuzgang@gmail.com>
2026-10-28 03:53:56 +08:00
..
2025-05-29 06:53:11 +00:00

RustFS

RustFS is a high-performance distributed object storage software built using Rust

CI Build and Push Docker Images GitHub commit activity Github Last Commit

Getting Started · Docs · Bug reports · Discussions

English | Simplified Chinese

RustFS is a high-performance distributed object storage software built using Rust, one of the most popular languages worldwide. Along with MinIO, it shares a range of advantages such as simplicity, broad S3 API compatibility for supported features, open-source nature, support for data lakes, AI, and big data. Furthermore, it has a better and more user-friendly open-source license in comparison to other storage systems, being constructed under the Apache license. As Rust serves as its foundation, RustFS provides faster speed and safer distributed features for high-performance object storage.

Features

  • High Performance: Built with Rust, ensuring speed and efficiency.
  • Distributed Architecture: Scalable and fault-tolerant design for large-scale deployments.
  • S3 Compatibility: Integration with common S3-compatible applications; current coverage is tracked in the S3 compatibility matrix.
  • Data Lake Support: Optimized for big data and AI workloads.
  • Open Source: Licensed under Apache 2.0, encouraging community contributions and transparency.
  • User-Friendly: Designed with simplicity in mind, making it easy to deploy and manage.

RustFS vs MinIO

Stress test server parameters

Type parameter Remark
CPU 2 Core Intel Xeon(Sapphire Rapids) Platinum 8475B , 2.7/3.2 GHz
Memory 4GB  
Network 15Gbp  
Driver 40GB x 4 IOPS 3800 / Driver

https://github.com/user-attachments/assets/2e4979b5-260c-4f2c-ac12-c87fd558072a

RustFS vs Other object storage

RustFS Other object storage
Powerful Console Simple and useless Console
Developed based on Rust language, memory is safer Developed in Go or C, with potential issues like memory GC/leaks
Does not report logs to third-party countries Reporting logs to other third countries may violate national security laws
Licensed under Apache, more business-friendly AGPL V3 License and other License, polluted open source and License traps, infringement of intellectual property rights
S3-compatible core, with coverage tracked in the compatibility matrix Variable S3 support and local cloud vendor coverage
Rust-based development, strong support for secure and innovative devices Poor support for edge gateways and secure innovative devices
Stable commercial prices, free community support High pricing, with costs up to $250,000 for 1PiB
No risk Intellectual property risks and risks of prohibited uses

Quickstart

To get started with RustFS, follow these steps:

  1. One-click installation script (Option 1)​​

    curl -O  https://rustfs.com/install_rustfs.sh && bash install_rustfs.sh
    
  2. Docker Quick Start (Option 2)​​

 # Docker Hub (recommended)
 docker run -d -p 9000:9000 -v /data:/data rustfs/rustfs:latest

 # Alternative using Podman
 podman run -d -p 9000:9000 -v /data:/data rustfs/rustfs:latest
  1. Access the Console: Open your web browser and navigate to http://localhost:9001 to access the RustFS console, default username and password is rustfsadmin .
  2. Create a Bucket: Use the console to create a new bucket for your objects.
  3. Upload Objects: You can upload files directly through the console or use S3-compatible APIs to interact with your RustFS instance.

Documentation

For detailed documentation, including configuration options, API references, and advanced usage, please visit our Documentation.

Getting Help

If you have any questions or need assistance, you can:

  • Check the FAQ for common issues and solutions.
  • Join our GitHub Discussions to ask questions and share your experiences.
  • Open an issue on our GitHub Issues page for bug reports or feature requests.

Contact

Contributors

RustFS is a community-driven project, and we appreciate all contributions. Check out the Contributors page to see the amazing people who have helped make RustFS better.

License

Apache 2.0

RustFS is a trademark of RustFS, Inc. All other trademarks are the property of their respective owners.