fix(storage): update seal_epoch synchronously

[zwang28]

I hereby agree to the terms of the Singularity Data, Inc. Contributor License Agreement.

What's changed and what's your intention?

Before this PR, seal_epoch updating is handled by HummockEventHandler asynchronously, which cannot ensure HummockEventHandler::seal_epoch is updated before notifying barrier finishing . It can lead to current epoch can't read, because the epoch in storage is not updated.

This PR fixes it by updating HummockStorage::seal_epoch synchronously, which shares the same underlying atomic var of HummockEventHandler::seal_epoch.

Checklist

• I have written necessary rustdoc comments
  - [ ] I have added necessary unit tests and integration tests
  - [ ] All checks passed in ./risedev check (or alias, ./risedev c)

Documentation

If your pull request contains user-facing changes, please specify the types of the changes, and create a release note. Otherwise, please feel free to remove this section.

Types of user-facing changes

Please keep the types that apply to your changes, and remove those that do not apply.

• Installation and deployment
• Connector (sources & sinks)
• SQL commands, functions, and operators
• RisingWave cluster configuration changes
• Other (please specify in the release note below)

Release note

Please create a release note for your changes. In the release note, focus on the impact on users, and mention the environment or conditions where the impact may occur.

Refer to a related PR or issue link (optional)

fix #7169

[wenym1]

Before this PR, the seal epoch should be updated in somewhere else. Should we remove updating the seal epoch there?

[hzxa21]

Before this PR, the seal epoch should be updated in somewhere else. Should we remove updating the seal epoch there?

+1. I think we don't need to update seal epoch when handling HummockEvent::SealEpoch in HummockEventHandler.

[zwang28]

Fixed.

[wenym1]

If this PR is intended to fix the panic caused by try_wait_epoch on uncommitted epoch, I would suggest we modify the logic of try_wait_epoch to do a really wait for the sealed_epoch to bump up asynchronously. In the future there may not be such global seal_epoch method, but instead seal each local state store, and then we may need to wait on the sealed_epoch of each local state store instance to bump up.

[codecov]

Codecov Report

Merging #7171 (d284a86) into main (0e90e6b) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main    #7171   +/-   ##
=======================================
  Coverage   73.15%   73.15%           
=======================================
  Files        1053     1053           
  Lines      167994   167996    +2     
=======================================
+ Hits       122891   122893    +2     
  Misses      45103    45103           

Flag	Coverage Δ
rust	73.15% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
...src/hummock/event_handler/hummock_event_handler.rs	80.53% <ø> (-0.05%)	⬇️
src/storage/src/hummock/state_store.rs	70.37% <100.00%> (+0.47%)	⬆️
src/meta/src/hummock/mock_hummock_meta_client.rs	64.73% <0.00%> (-1.58%)	⬇️
...torage/src/hummock/local_version/pinned_version.rs	88.30% <0.00%> (-0.59%)	⬇️
src/object_store/src/object/mem.rs	87.12% <0.00%> (-0.38%)	⬇️
src/storage/src/hummock/compactor/mod.rs	83.10% <0.00%> (-0.31%)	⬇️
src/object_store/src/object/mod.rs	51.08% <0.00%> (-0.22%)	⬇️
src/storage/src/hummock/sstable_store.rs	63.58% <0.00%> (-0.20%)	⬇️
src/meta/src/hummock/manager/mod.rs	79.33% <0.00%> (-0.12%)	⬇️
src/batch/src/executor/group_top_n.rs	74.85% <0.00%> (+6.43%)	⬆️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[wenym1]

Anyway, the current fix LGTM. We may introduce the real wait in future PR.

[zwang28]

I would suggest we modify the logic of try_wait_epoch to do a really wait for the sealed_epoch to bump up asynchronously

Agree.
I'll merge this quick fix first, to avoid flaky tests.

[hzxa21]

LGTM