fix(sentinel): preserve root seeds for outage recovery

[nkaradzhov]

fixes #3127

Description

Describe your pull request here

---

Checklist

• Does npm test pass with this change (including linting)?
• Is the new or changed code fully tested?
• Is a documentation update included (if this change modifies existing APIs, or introduces new ones)?

---

Note

Medium Risk
Changes sentinel failover/rediscovery behavior during disconnects by repopulating #sentinelRootNodes from stored seed nodes, which could affect reconnection loops and topology updates under real outages. Added coverage reduces risk but behavior is timing-sensitive and exercised during network partitions.

Overview
Improves Sentinel outage recovery by persisting the initially configured sentinel endpoints (#sentinelSeedNodes) and restoring #sentinelRootNodes from those seeds whenever failures remove all known roots (also enforced at the start of observe()).

Makes sentinel list change detection compare the actual node sets (via a stable host/port key) instead of just list length, and adds an integration test that simulates a full Redis+Sentinel outage and asserts periodic commands recover after restart.

^{Written by Cursor Bugbot for commit d8421b0. This will update automatically on new commits. Configure here.}