chore: ensure there is no staging data in state table on vnode bitmap update

[hzxa21]

I hereby agree to the terms of the RisingWave Labs, Inc. Contributor License Agreement.

What's changed and what's your intention?

Prior to this PR, we only check whether memtable is empty in local state store on vnode bitmap update. This PR adds another assertion to ensure there is no staging data in local state store (i.e. all data must be committed via checkpoiont) to ensure that no stale data will be read after scaling.

Checklist

• I have written necessary rustdoc comments
• I have added necessary unit tests and integration tests
• I have added test labels as necessary. See details.
• I have added fuzzing tests or opened an issue to track them. (Optional, recommended for new SQL features Sqlsmith: Sql feature generation #7934).
• My PR contains breaking changes. (If it deprecates some features, please create a tracking issue to remove them in the future).
• All checks passed in ./risedev check (or alias, ./risedev c)
• My PR changes performance-critical code. (Please run macro/micro-benchmarks and show the results.)

• My PR contains critical fixes that are necessary to be merged into the latest release. (Please check out the details)

Documentation

• My PR needs documentation updates. (Please use the Release note section below to summarize the impact on users)

Release note

If this PR includes changes that directly affect users or other significant modifications relevant to the community, kindly draft a release note to provide a concise summary of these changes. Please prioritize highlighting the impact these changes will have on users.

[wenym1]

I think a cleaner way is to let local state store be aware of its own vnode bitmap. We can pass the initial vnode bitmap in init via InitOption, and then we can let it be aware of vnode bitmap update in like seal_current_epoch. When it knows it's going to update the vnode bitmap, it then asserts that there is no staging data. IIRC, all call on update_vnode_bitmap of StateTable is beside commit.

[hzxa21]

I think a cleaner way is to let local state store be aware of its own vnode bitmap. We can pass the initial vnode bitmap in init via InitOption, and then we can let it be aware of vnode bitmap update in like seal_current_epoch. When it knows it's going to update the vnode bitmap, it then asserts that there is no staging data. IIRC, all call on update_vnode_bitmap of StateTable is beside commit.

Actually I am proposing this PR because I am writing a PR to let local state store be aware of vnode bitmap. It is indeed cleaner to do the assertion when local state store is updating the vnode bitmap. However, I am still thinking whether we should do vnode bitmap update in seal_current_epoch because I saw some logics exist in between update_vnode_bitmap and commit for some executors.

[wenym1]

because I saw some logics exist in between update_vnode_bitmap and commit for some executors.

We can decide whether to remove this method in the future after we know whether we can move these logic to elsewhere and make update_vnode_bitmap and commit atomic.

[hzxa21]

#15183