refactor(logging): standardize rustfs runtime events (#3396)

* fix(rpc): adjust log levels and reduce noise in storage RPC layer

Issue #682: pool rebalance normal flow logs were using warn! instead
of info!. Additionally, several high-frequency RPC entry points (ping,
write_stream, read_at, walk_dir) were logging at info! level, causing
unnecessary log noise in production.

Changes:
- load_rebalance_meta: warn! → info! for normal flow (3 occurrences)
- ping body decode: info! → debug! (fires on every health check)
- write_stream/read_at/walk_dir entry: info! → debug!
- metrics.rs: lowercase error messages, add structured error field
- http_service.rs: add structured error field to walk_dir failure
- Add tracing::instrument fields for start_rebalance context

* refactor(logging): standardize rustfs runtime events

* build(deps): bump workspace dependency versions

* build(deps): pin time to 0.3.47

* build(deps): update postgres client versions

crates/e2e_test/Cargo.toml

@@ -67,7 +67,7 @@ chrono = { workspace = true }
 md5 = { workspace = true }
 sha2 = { workspace = true }
 astral-tokio-tar = { workspace = true }
-s3s.workspace = true
+s3s = { workspace = true }
 zstd.workspace = true
 time.workspace = true
 suppaftp = { workspace = true, features = ["tokio", "rustls-aws-lc-rs"] }