fix bug: issue #3595 and add unit test

[kael-aiur]

Make sure that:

• You have read the contribution guidelines.
• You have created a feature request first to discuss your contribution intent. Please reference the feature request ticket number in the pull request.
• You applied code formatting rules using the mvn formatter:format target. Don’t submit any formatting related changes.
• You submit test cases (unit or integration tests) that back your changes.

---

Note

Medium Risk
Changes stale-connection detection used during partition reconfiguration, which can affect connection lifecycle and traffic routing during failover. Scope is small and covered by a new unit test, but incorrect role detection could cause extra reconnects or missed closures.

Overview
Fixes stale-connection detection during cluster partition updates so cached connections are treated as stale when the node’s current role no longer matches the connection intent (e.g., a READ connection pointing at an upstream/master, or a WRITE connection pointing at a replica).

Adds a unit test ensuring a cached read-from-replica connection is closed when that replica is promoted to master after setPartitions().

^{Written by Cursor Bugbot for commit 160ebad. This will update automatically on new commits. Configure here.}

[a-TODO-rov]

Thanks @kael-aiur
Will review this soon.

[jit-ci]

🛡️ Jit Security Scan Results

✅ No security findings were detected in this PR

---

^{Security scan by Jit}

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, have a team admin enable autofix in the Cursor dashboard.}

[kael-aiur]

I think the test fail on redis 7.4 and 8.2 is appears to be a flaky test unrelated to my changes. The test failure is in scriptLoad() which is unrelated to the clusterMyId() fallback logic I modified？