What our run establishes
Our run on August 24, 2026, checked commit 3b0a28d in the lab-rust:1 sandbox. Our build and test commands each timed out after 900 seconds. Installation completed in 43 seconds, and the stored dependency count is 1,222.
The harness records each command's outcome, elapsed time, exit code and timeout flag separately. Its overall build_failed status must be read alongside those details: both unfinished commands have exit code 124 and timedOut set to true. The build has no retained output tail. The test tail ends with compilation output rather than a test-result footer, so it does not establish whether any assertions ran elsewhere in the log. Parsed test counts are absent.
This was a source setup attempt, not a storage workload trial. We did not verify a running server, Docker deployment, S3 client compatibility, object durability, node-loss recovery, KMS restart recovery, upgrades or throughput. The audit result is absent, and scan coverage is unknown. The testing methodology explains the publisher's command-outcome reporting and limits; its current text does not establish which methodology version governed this older run. No new execution underlies this correction.
For contributors, the useful result is a budget constraint: this sandbox did not complete either command within its limit. A longer run or a different build environment would be needed to determine completion. For storage buyers, the next useful step is an acceptance trial against a pinned deployment, with recoverable data and explicit failure tests.
Current feature status differs from the tested checkout
RustFS documents S3 compatibility for supported features and an Apache 2.0 license. Those are reasons to investigate it, while compatibility with your application's operations remains something to verify.
The README at our recorded commit marked distributed mode, lifecycle management and RustFS KMS as under testing. The supplied current README marks all three available. That is a change in the project's stated status, not a result from our sandbox. The current README defines available features as shipped and covered by its CI gates; we did not independently run those gates.
The current README also draws boundaries that matter to deployment planning. Swift and SFTP require opt-in build features. MinIO on-disk compatibility is a preview behind rio-v2, outside the default build, and MinIO-encrypted objects are not readable by RustFS. A single-node single-drive deployment cannot expand in place or become a pool; moving to a multi-drive topology requires a new deployment and S3 migration. Consult the RustFS README before choosing the initial layout.
The listing report was corrected and closed
Issue #6218 initially described listed keys whose HEAD and GET requests returned 404. Later investigation changed the interpretation of the captured examples. On August 29, the reporter confirmed that version-qualified reads returned the complete object bodies. The problem involved a stale per-node version stack after delete markers reached only part of the cluster, rather than demonstrated loss of those captured objects.
The issue was closed through #6977 on August 31. In the supplied September 1 follow-up, the reporter still planned field validation on a build containing the fix. Closure records an upstream resolution; that extract does not establish the outcome of the planned validation. A useful acceptance test therefore compares flat listings with reads during lifecycle deletion and rolling maintenance, while preserving version IDs and per-node evidence when results disagree.
The supplied readiness report, #6286, was also closed after a fix, with a maintainer requesting a retest. It should not be presented as an unresolved defect in the current project. We did not reproduce it or verify its resolution. Likewise, no independently fetched extract for the earlier KMS restart report was supplied, so this review does not repeat it as an established current limitation.
Alternatives depend on the job
SeaweedFS documents S3 object storage and a POSIX file system over the same data. Evaluate it as a replacement candidate when shared file and object access is a requirement. That is a documented capability, not evidence that it will outperform RustFS or meet your durability requirements. Its README provides the deployment paths to investigate.
The MinIO repository states that it is no longer maintained. Its README points readers to AIStor Free for a standalone edition and AIStor Enterprise for a distributed edition with commercial support. Treat those as separate product evaluations; the old community repository is not a default recommendation for a newly maintained deployment. MinIO's client tooling can also be a complement to an S3 evaluation rather than a replacement storage server.
How to read the retained scores
The numerical scores below retain the recorded assessment. The rubric is 1 for substantial documented obstacles, 2 for material constraints requiring evaluation, 3 for adequate evidence for the stated scope, 4 for detailed supporting evidence with clear limits, and 5 for strong evidence including independent verification in the relevant scope. These are editorial judgments, not benchmark results or production certifications.
Setup ease is 2 because neither source command completed within the sandbox limit, while Docker remains a documented but untested route. Documentation is 4 for explicit mount-permission instructions, feature boundaries and topology warnings. Community is 4 for the supplied issue discussions' detailed artifacts, technical responses and follow-through; it does not measure adoption or support response guarantees. Maturity remains the recorded 2 for the evaluated release-candidate scope, where distributed operation was marked under testing and our runtime validation was absent. It is not a fresh numerical assessment of the current release.

