robustness: only run MemberDowngrade test for high SnapshotCatchUpEntries

[siyuanfoundation]

This should prevent snapshot from being sent in the test.
Tested locally for 100 times, did not see any flakiness.

#19147

Please read https://github.com/etcd-io/etcd/blob/main/CONTRIBUTING.md#contribution-flow.

[k8s-ci-robot]

Skipping CI for Draft Pull Request.
If you want CI signal for your change, please convert it to an actual PR.
You can still manually trigger a test run with /test all

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 68.80%. Comparing base (0edc5c0) to head (08e4d6d).
Report is 2 commits behind head on main.

Additional details and impacted files

see 18 files with indirect coverage changes

@@           Coverage Diff           @@
##             main   #19160   +/-   ##
=======================================
  Coverage   68.79%   68.80%           
=======================================
  Files         420      420           
  Lines       35649    35649           
=======================================
+ Hits        24526    24527    +1     
+ Misses       9695     9692    -3     
- Partials     1428     1430    +2     

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 0edc5c0...08e4d6d. Read the comment docs.

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: serathius, siyuanfoundation

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [serathius]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[serathius]

Awesome finding! Thanks @siyuanfoundation

[ahrtr]

Nice, it's a better solution. Each member will have complete WAL records.

But theoretically, it's still possible that a follower may lag behind the leader more than the DefaultSnapshotCatchUpEntries in high traffic case, so the leader will still send a snapshot, although it's low possibility.

Tested locally for 100 times, did not see any flakiness.

The other reason why you do not see the error is 50% of downgrade cases out of the 100 tries were skipped due to the catchUpEntries too small.

[serathius]

The other reason why you do not see the error is 50% of downgrade cases out of the 100 tries were skipped due to the catchUpEntries too small.

My assumption it was 100 runs without the skip

[ahrtr]

My assumption it was 100 runs without the skip

In this PR, it set two snapCatchUpEntries (100, etcdserver.DefaultSnapshotCatchUpEntries), while robustness randomly selects one of them, so it's 50% possibilities? Once 100 is selected, then the downgrade test is skipped.

Anyway, it's already good enough. Up to your guys to decide whether & how to enhance it.