[Bug]: pkg/logservice.TestAddHeartbeat times out in base CI

### Is there an existing issue for the same bug?

- [x] I checked the open and closed issue search.

The restore timeout from the same base-only CI job is tracked separately by #28763. This issue is only for the independent `pkg/logservice.TestAddHeartbeat` Dragonboat timeout signature.

### CI evidence

- Source PR: #28884; the PR changes only a MongoDB lease-test helper.
- Workflow run: https://github.com/matrixorigin/matrixone/actions/runs/34921783776
- Failed job: https://github.com/matrixorigin/matrixone/actions/runs/34921783776/job/104231337338
- PR head: `564d644377024349fc441126ab37987059e9a18d`
- Runner: Ubuntu/x86 UT job.
- Exact failure: `github.com/matrixorigin/matrixone/pkg/logservice.TestAddHeartbeat`, after 0.68s; `store.addLogStoreHeartbeat` returned Dragonboat `timeout` at `pkg/logservice/store_test.go:705`.

### Assessment

This is base-owned CI evidence, not a regression in #28884: the PR has no changes in the logservice/HAKeeper heartbeat path and its head already contains current `main`. A single occurrence does not establish a safe production fix. It is distinct from the restore timeout class tracked by #28763 and from the historical HAKeeper bootstrap race in #28550.

### Queue / acceptance

Assigned to `XuPeng-SH` for CI-load/readiness triage. Reproduce or correlate this signature with HAKeeper/Dragonboat readiness and resource diagnostics first; then queue a focused fix PR only after a deterministic root cause and a bounded regression are established. Do not mask the failure with retries, sleeps, or a blanket timeout increase. No speculative fix PR is opened from this one base-only occurrence.