fix(ledger): base replay slot has to be next slot after last persisted one

[taco-paco]

Summary

Base replay slot has to be next slot after last persisted one

Breaking Changes

• None
• Yes — migration path described below

Test Plan

Summary by CodeRabbit

Bug Fixes

• Adjusted ledger synchronization behavior to correct the processing start position when the database falls behind, ensuring proper consistency during ledger replay operations.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro

Run ID: a7e1f1c0-847a-478b-a247-a59e07c60d83

📥 Commits

Reviewing files that changed from the base of the PR and between f001043 and af6af0c.

📒 Files selected for processing (1)

• magicblock-api/src/magic_validator.rs

---

📝 Walkthrough

Walkthrough

This PR adjusts the ledger replay logic in magicblock-api/src/magic_validator.rs. When the accounts database lags behind the ledger, the replay process now begins at the slot immediately following the accounts database slot (accountsdb_slot + 1) rather than at the accounts database slot itself. The change introduces a new variable full_process_starting_slot to compute and pass this adjusted boundary into the process_ledger function.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/replay-base-slot

---

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.}