KVStore: fix tiflash_raft_throughput_bytes warp problem

[CalvinNeo]

What problem does this PR solve?

Issue Number: close #10701

Problem Summary:

What is changed and how it works?

Check List

Tests

• Unit test
• Integration test
• Manual test (add detailed scripts or steps below)
• No code

Side effects

• Performance regression: Consumes more CPU
• Performance regression: Consumes more Memory
• Breaking backward compatibility

Documentation

• Affects user behaviors
• Contains syntax changes
• Contains variable changes
• Contains experimental features
• Changes MySQL compatibility

Release note

Fixed an issue where Grafana's Raft throughput might incorrectly display very large values.

Summary by CodeRabbit

• Bug Fixes
  • Safer storage size accounting implemented to prevent underflow/overflow on edge-case writes.
  • Size updates unified across insert paths, improving accuracy of write-size reporting and reducing risk of rare errors or crashes while preserving existing insert behavior.

[coderabbitai]

📝 Walkthrough

Walkthrough

Adds a guarded local updater in RaftCommands.cpp that centralizes write_size adjustments: entries compute a payload (unifying v2 and non-v2 paths), then call update_write_size(payload) which handles positive, negative, and INT64_MIN edge cases to avoid underflow.

Changes

Cohort / File(s)	Summary
Raft Payload Handling dbms/src/Storages/KVStore/MultiRaft/RaftCommands.cpp	Replaced direct write_size modifications with a local update_write_size(payload) lambda; both v2 and non-v2 insert branches now compute a single payload via doInsert/dup-check and call the updater, which clamps negative adjustments and handles INT64_MIN edge case to prevent underflow.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~8 minutes

Poem

🐇 I nibble bytes with gentle paws,
Tallying payloads without a pause,
Clamp the dips, dodge the fall,
Keep counts steady, one and all —
A cautious hop, a guarded cause.

🚥 Pre-merge checks | ✅ 2 | ❌ 3

❌ Failed checks (2 warnings, 1 inconclusive)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The PR description follows the template structure but has critical gaps: the 'What is changed and how it works?' section is empty, no commit message is provided, and no tests are documented.	Fill in the 'What is changed and how it works?' section with implementation details, provide commit message, and specify what tests were performed.
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.
Linked Issues check	❓ Inconclusive	Issue #10701 is incomplete—it lacks reproduction steps, expected vs. actual outputs, and version details needed to validate whether the code changes address the reported metric warp problem.	Complete issue #10701 with reproduction steps, expected behavior, and actual behavior observed to confirm the fix addresses the root cause.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title 'KVStore: fix tiflash_raft_throughput_bytes warp problem' is specific and clearly indicates the module and the bug being fixed.
Out of Scope Changes check	✅ Passed	Code changes focus on safe write_size handling and payload accounting in RaftCommands, which aligns with fixing a metric warp issue; no unrelated modifications detected.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 1

🤖 Fix all issues with AI agents

In `@dbms/src/Storages/KVStore/MultiRaft/RaftCommands.cpp`:
- Around line 385-398: The code mixes signed payload with unsigned write_size
causing wrap/overflow; fix by treating payload as a signed integer before doing
arithmetic: capture payload from doInsert into a signed type (e.g., long long)
such as payload_signed = static_cast<long long>(... .payload), then if
(payload_signed > 0) do write_size += static_cast<size_t>(payload_signed); else
if (write_size >= static_cast<size_t>(-payload_signed)) do write_size -=
static_cast<size_t>(-payload_signed). Apply this change in both branches that
call doInsert (the blocks guarded by is_v2 and the else), referencing doInsert,
payload, write_size and DupCheck.

[ti-chi-bot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: JaySon-Huang, JinheLin

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

The pull request process is described here

Details

Needs approval from an approver in each of these files:

• OWNERS [JaySon-Huang,JinheLin]

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

[ti-chi-bot]

[LGTM Timeline notifier]

Timeline:

• 2026-02-05 03:55:45.858370432 +0000 UTC m=+325616.959769150: ☑️ agreed by JaySon-Huang.
• 2026-02-05 07:18:23.950297304 +0000 UTC m=+337775.051696020: ☑️ agreed by JinheLin.

[JaySon-Huang]

/cherry-pick release-8.5

[ti-chi-bot]

@JaySon-Huang: new pull request created to branch release-8.5: #10705.
But this PR has conflicts, please resolve them!

Details

In response to this:

/cherry-pick release-8.5

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the ti-community-infra/tichi repository.