fix: retry disk writes dropped during ReplaceComponentList resync window to prevent silent object loss in edge cache

[WorrierKhushal]

What type of PR is this?

/kind bug

What this PR does / why we need it:

During a cloud-reconnect resync, ReplaceComponentList() in
pkg/yurthub/storage/disk/storage.go holds a pending lock
(lockComponentKey()) across its full disk I/O sequence. Any concurrent
watch events arriving for objects under the same key prefix during this
window hit ErrStorageAccessConflict in Update()/Create() and are
permanently dropped from disk.

PR #2712 correctly fixed the in-memory divergence half of this problem —
it prevented the in-memory cache from getting ahead of disk when a
conflict occurs. But the disk write itself was still never retried, so
the on-disk snapshot after every resync silently omits any objects that
arrived via watch during the replace window.

On the next edge-autonomous period (cloud disconnect), YurtHub serves
this incomplete snapshot to Kubelet — missing objects that were
legitimately created or updated just before the disconnect. The missing
state is permanent on disk until the next full resync.

This PR fixes the second half: when storeObjectWithKey() catches
ErrStorageAccessConflict, it now enqueues the key+object into a small
pendingWrites buffer on CacheManager. After ReplaceComponentList()
completes and releases its lock, saveListObject() drains this buffer
using the existing saveOneObject() write path — reusing already-tested
infrastructure, no new disk I/O patterns, no storage.Store interface
changes.

Added a deterministic regression test that forces the exact conflict
window using a channel-controlled fake StorageWrapper (same pattern
as the test added for #2712), confirms the object is absent from disk
before the fix, and present after.

Which issue(s) this PR fixes:

Fixes #2780

Special notes for your reviewer:

• go test -race -v ./pkg/yurthub/cachemanager/... — zero races, all
  tests pass including the new regression test
• go build and go vet on both cachemanager and storage/disk —
  clean
• storage.Store interface unchanged — no downstream implementers
  affected
• saveOneObject() unchanged — reused as the drain write path only
• This completes the fix that [BUG] Silent watch event drop during disk storage list-replacement causes permanent cache inconsistency between memory and disk #2712 started — that PR fixed memory/disk
  divergence; this PR fixes the disk-only silent object loss

Does this PR introduce a user-facing change?

fix(yurthub): prevent silent loss of watch-event objects from disk cache 
during cloud-reconnect resync window; objects that arrived via watch 
while ReplaceComponentList held its lock are now reliably written to disk 
after the replace completes.

other Note

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[codecov]

Codecov Report

❌ Patch coverage is 76.19048% with 5 lines in your changes missing coverage. Please review.
✅ Project coverage is 46.29%. Comparing base (ddf22f7) to head (b95999a).
⚠️ Report is 7 commits behind head on master.

Files with missing lines	Patch %	Lines
pkg/yurthub/cachemanager/cache_manager.go	75.00%	3 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #2786      +/-   ##
==========================================
+ Coverage   46.20%   46.29%   +0.09%     
==========================================
  Files         405      405              
  Lines       27540    27583      +43     
==========================================
+ Hits        12724    12769      +45     
+ Misses      13649    13636      -13     
- Partials     1167     1178      +11     

Flag	Coverage Δ
unittests	46.29% <76.19%> (+0.09%)	⬆️

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.