[fix](snapshot) drop index should set db_id (#58401)

[w41ter]

cherry-pick #58401

recycle_indexes require db_id to recycle multi versioned keys, but drop index does not set db_id

[Thearas]

Thank you for your contribution to Apache Doris.
Don't know what should be done next? See How to process your PR.

Please clearly describe your PR:

1. What problem was fixed (it's best to include specific error reporting information). How it was fixed.
2. Which behaviors were modified. What was the previous behavior, what is it now, why was it modified, and what possible impacts might there be.
3. What features were added. Why was this function added?
4. Which code was refactored and why was this part of the code refactored?
5. Which functions were optimized and what is the difference before and after the optimization?

[w41ter]

run buildall

[w41ter]

/review

[github-actions]

No blocking issues found.

Critical checkpoints:

• Correctness: Pass. The patch consistently threads through the FE call chain that reaches cloud , including table erase, rollup/schema-change cancel, MV/rollup drop, restore, recycle-bin cleanup, and replay.
• Completeness: Pass. I did not find remaining FE-side callers of , , , or in this branch that still omit .
• Compatibility: Pass for the touched paths. The replay/recycle metadata used here already carries (, ), so the new argument is available where this PR consumes it.
• Tests: No new automated coverage was added in this PR. Existing cloud/meta-service tests already exercise with , but FE-side propagation still relies on code inspection rather than a dedicated regression test.

User focus:

• No additional user-provided review focus was supplied.

Residual risk:

• I did not run FE unit tests in this review session.

[github-actions]

No blocking issues found.

Critical checkpoints:

• Correctness: Pass. The patch consistently threads dbId through the FE call chain that reaches cloud drop_index, including table erase, rollup/schema-change cancel, MV/rollup drop, restore, recycle-bin cleanup, and replay.
• Completeness: Pass. I did not find remaining FE-side callers of onEraseOlapTable, eraseTableDropBackendReplicas, eraseDroppedIndex, or dropMaterializedIndex in this branch that still omit dbId.
• Compatibility: Pass for the touched paths. The replay/recycle metadata used here already carries dbId (DropInfo, RecycleTableInfo), so the new argument is available where this PR consumes it.
• Tests: No new automated coverage was added in this PR. Existing cloud/meta-service tests already exercise drop_index with db_id, but FE-side propagation still relies on code inspection rather than a dedicated regression test.

User focus:

• No additional user-provided review focus was supplied.

Residual risk:

• I did not run FE unit tests in this review session.

[hello-stephen]

FE UT Coverage Report

Increment line coverage 37.50% (6/16) 🎉
Increment coverage report
Complete coverage report

[hello-stephen]

FE Regression Coverage Report

Increment line coverage 75.00% (12/16) 🎉
Increment coverage report
Complete coverage report

[w41ter]

run buildall

[hello-stephen]

FE Regression Coverage Report

Increment line coverage 75.00% (12/16) 🎉
Increment coverage report
Complete coverage report

[morningman]

run buildall

[hello-stephen]

FE UT Coverage Report

Increment line coverage 37.50% (6/16) 🎉
Increment coverage report
Complete coverage report

[hello-stephen]

FE Regression Coverage Report

Increment line coverage 87.50% (14/16) 🎉
Increment coverage report
Complete coverage report

[hello-stephen]

FE Regression Coverage Report

Increment line coverage 4.88% (14/287) 🎉
Increment coverage report
Complete coverage report

[hello-stephen]

FE Regression Coverage Report

Increment line coverage 3.46% (14/405) 🎉
Increment coverage report
Complete coverage report