Commit Blocks in Snapshot and Enhance UT

[yuwmao]

Commit Blocks in snapshot, or the blocks will be overwritten after restart.
Add more UTs to test logic when follower/leader restart.

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

Attention: Patch coverage is 15.38462% with 11 lines in your changes missing coverage. Please review.

Project coverage is 62.46%. Comparing base (1746bcc) to head (d177b4c).
Report is 37 commits behind head on main.

Files with missing lines	Patch %	Lines
...ib/homestore_backend/replication_state_machine.cpp	0.00%	7 Missing ⚠️
...lib/homestore_backend/snapshot_receive_handler.cpp	20.00%	3 Missing and 1 partial ⚠️

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #258      +/-   ##
==========================================
- Coverage   63.15%   62.46%   -0.70%     
==========================================
  Files          32       33       +1     
  Lines        1900     2587     +687     
  Branches      204      307     +103     
==========================================
+ Hits         1200     1616     +416     
- Misses        600      808     +208     
- Partials      100      163      +63     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[xiaoxichen]

lgtm, a bit concern on the leader we assigned during creation might not be the leader when we doing the test (if leader switch happened in between). But as far as this is testing code, it 's LGTM

[yuwmao]

lgtm, a bit concern on the leader we assigned during creation might not be the leader when we doing the test (if leader switch happened in between). But as far as this is testing code, it 's LGTM

Yes, it could occur, will improve it by using get_leader_id instead of constant initial_leader_replica_id as far as possible to reduce the chances of this case occurring.

[xiaoxichen]

lgtm, a bit concern on the leader we assigned during creation might not be the leader when we doing the test (if leader switch happened in between). But as far as this is testing code, it 's LGTM

Yes, it could occur, will improve it by using get_leader_id instead of constant initial_leader_replica_id as far as possible to reduce the chances of this case occurring.

But if the leader is replica-0 it is hard to handle :). Maybe explicity request leadership is better . Anyway lets improve later, feel free to merge as is

[yuwmao]

lgtm, a bit concern on the leader we assigned during creation might not be the leader when we doing the test (if leader switch happened in between). But as far as this is testing code, it 's LGTM

Yes, it could occur, will improve it by using get_leader_id instead of constant initial_leader_replica_id as far as possible to reduce the chances of this case occurring.

But if the leader is replica-0 it is hard to handle :). Maybe explicity request leadership is better . Anyway lets improve later, feel free to merge as is

Ok, if it supports to force change leadership to the specified member, that will be easy, we can request and wait for leader.