mirror of
https://github.com/rustfs/rustfs.git
synced 2026-07-26 16:28:15 +00:00
weisd
9dcf8cb7ea
* fix: unblock empty chunked admin rebalance start Some admin clients send empty POST requests with chunked framing and no explicit content length. The admin rebalance route carries no request body, so normalize that known route before downstream validation sees a missing length. Constraint: Keep normalization scoped to known empty-body admin routes Rejected: Relax transfer-encoding handling for S3 routes | broader protocol risk Confidence: high Scope-risk: narrow Tested: cargo test -p rustfs --lib server::layer::tests:: Tested: cargo fmt --all --check Tested: git diff --check * fix: keep rebalance listing from stalling Large object migration was awaited inside the listing callback, so metacache readers could stop being drained while drive walk operations were still writing listing data. Move entry migration into bounded background tasks and wait for them after each set listing completes. Constraint: Preserve existing rebalance cancellation and first-error propagation Rejected: Only increase drive walk timeouts | masks callback backpressure and still fails for slower migrations Confidence: medium Scope-risk: moderate Tested: cargo test -p rustfs-ecstore rebalance_unit_tests Tested: cargo test -p rustfs-ecstore cache_value::metacache_set::tests Tested: cargo test -p rustfs --lib server::layer::tests:: Tested: cargo clippy -p rustfs-ecstore --lib --all-features -- -D warnings Tested: cargo clippy -p rustfs-ecstore --tests --all-features -- -D warnings Tested: cargo fmt --all --check Tested: git diff --check --------- Co-authored-by: loverustfs <hello@rustfs.com> Co-authored-by: houseme <housemecn@gmail.com> Co-authored-by: cxymds <Cxymds@qq.com>