Performance improvements and guide update

[tishun]

Summary

Revamps the performance/benchmark suite, lands two structural performance improvements to the core locking paths (parallel multi-node I/O and configurable exponential backoff with jitter), brings the documentation back in line with the current API, and streamlines CI. Core code changes are additive and non-breaking.

Core performance

• Parallel multi-node I/O (MultiNodeStrategy): acquire/release/extend now fan out per-node Redis operations concurrently via CompletableFuture on a shared daemon thread pool, waiting for the quorum on join. Per-attempt cost drops from ~N×RTT to ~max(RTT) on multi-node clusters.
• Exponential backoff with jitter: new BackoffCalculator (delay = min(maxRetryDelay, retryDelay × multiplier^attempt), optional jitter), applied in PollingWaitStrategy, the Redlock retry-loop fallback, and RedlockReadWriteLock. Replaces the fixed retry delay that caused contention storms. FairLock intentionally stays on tight polling.
• Config: RedlockConfiguration adds maxRetryDelay, retryDelayMultiplier (≥ 1.0), and retryDelayJitterRatio ([0.0, 1.0]), validated. Defaults preserve prior behavior (multiplier 1.0, jitter 0.0).
• LockWaitStrategy: attempt-aware waitForRelease(key, timeout, attempt) and a backoff-parameterized initialize(...) overload (both default-provided).

Benchmarks & tests

• New JsonReportGenerator emits machine-readable JSON alongside markdown; per- primitive *-benchmark-results.json added; scenarios/clients and aggregation refactored; new benchmark-analysis.md with methodology and cross-library standings.
• Added BackoffCalculatorTest, PollingWaitStrategyTest, and new RedlockConfigurationTest cases.

Documentation

• Full accuracy pass against the current API. Rewrote the getting-started path (index, quick-start, basic-usage, redis-clients, best-practices) off the removed legacy new Redlock(pool…) / redlock.lock(key, ttl) API to the
  RedlockManager factory (withJedis/withLettuce → createLock, JDK Lock, try-with-resources on the manager).
• API reference: fixed RxRedlock method names/return types, documented the new backoff config options and database node option.
• Architecture: corrected native CAS/CAD to Redis 8.4+, clarified polling uses the configurable retryDelay (default 200ms); regenerated stale benchmark tables to match the current results.
• Comparison docs: switched to factory usage, corrected CountDownLatch.countDown to atomic (decrAndPublishIfZero), fixed signatures, removed leftover fragments.
• Removed decorative emoji across the docs.

CI

• ci.yml: consolidated unit + integration test steps, moved performance tests to -Dgroups=performance, and removed redundant report/code-quality/build-info jobs.
• Removed the standalone pr-validation.yml workflow.
• nightly.yml: dropped a redundant unit-test step.
• Added a Codecov coverage badge to the README.

Compatibility

No breaking API changes. New config options are additive; new interface methods have defaults. Behavior is unchanged unless backoff is explicitly configured.

[codecov]

Codecov Report

❌ Patch coverage is 85.18519% with 16 lines in your changes missing coverage. Please review.
✅ Project coverage is 74.07%. Comparing base (371dfe9) to head (2d85413).
⚠️ Report is 4 commits behind head on main.

Files with missing lines	Patch %	Lines
...codarama/redlock4j/strategy/BackoffCalculator.java	66.66%	2 Missing and 3 partials ⚠️
...codarama/redlock4j/strategy/MultiNodeStrategy.java	82.14%	5 Missing ⚠️
...n/java/org/codarama/redlock4j/AbstractRedlock.java	62.50%	2 Missing and 1 partial ⚠️
...a/org/codarama/redlock4j/RedlockReadWriteLock.java	85.71%	0 Missing and 1 partial ⚠️
.../redlock4j/configuration/RedlockConfiguration.java	95.23%	0 Missing and 1 partial ⚠️
...darama/redlock4j/strategy/PollingWaitStrategy.java	91.66%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@              Coverage Diff              @@
##               main      #20       +/-   ##
=============================================
+ Coverage     61.21%   74.07%   +12.86%     
=============================================
  Files            25       27        +2     
  Lines          1748     1809       +61     
  Branches        200      208        +8     
=============================================
+ Hits           1070     1340      +270     
+ Misses          597      341      -256     
- Partials         81      128       +47     

Flag	Coverage Δ
unittests	74.07% <85.18%> (+12.86%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.