feat: mechanism to calculate last settled block by not using `metadata` #838

goran-ethernal · 2025-08-07T09:56:47Z

🔄 Changes Summary

This PR introduces a new interface called CertificateQuerier which, for now, has only one function called GetLastSettledCertificateToBlock which calculates the last settled block from a previously settled certificate.

This will be used on the validator, to calculate what was the last settled block, so we can figure out from which block a new certificate that is being verified is built.

It will also be used on startup from scratch, when aggsender syncs its state with agglayer, so we can save the ToBlock of the last settled certificate in the db.

The motivation behind this PR is to not use the Metadata field on the certificate, since it will not be signed by the proposer and validators, hence, it can be faked, but deduce the block range from other certificate data.

How the last settled block is calculated

The last settled block in the CertificateQuerier is calculated like this:

First it will check the NewLocalExitRoot field on the certificate, if it is not the first empty LER, meaning, there were some bridge exits in the certificate, or in certificates before it, it will get the block num on which the NewLocalExitRoot hash was added on L2 by calling the L2 bridge syncer.
Secondly, it will call the agglayer endpoint to get the latest settled imported bridge exit. Note that this PR added only the endpoint sceleton, since this is not yet implemented on agglayer. This endpoint will return the GlobalIndex of the last settled imported bridge exit (claim), and we will use that to get the claim from the L2 bridge syncer since GlobalIndex is unique per claim.
Thirdly, it will ping the AggchainFEP rollup contract to get the last settled L2 block from it (which is gotten from the aggchain proof in the FEP certificates). If the network is not FEP this call will just return 0.
At the end, it will compare the 3 blocks gotten in previous steps, and choose the max value from it.

We need all three values because, PP certificates can either contain only bridge exits, or only imported bridge exits, and FEP certificates can be empty.

IMPORTANT NOTE: This code is not used anywhere. It will be used in subsequent PRs.

⚠️ Breaking Changes

NA

📋 Config Updates

NA

✅ Testing

🤖 Automatic: aggkit CI

Copilot

Pull Request Overview

This PR introduces a CertificateQuerier interface and implementation to calculate the last settled block from previously settled certificates without relying on the Metadata field. The implementation examines three sources of settlement data: bridge exits via NewLocalExitRoot, imported bridge exits from the agglayer, and FEP contract settlement data.

Key changes:

New CertificateQuerier interface with GetLastSettledCertificateToBlock method
AggchainFEPRollupQuerier interface and implementation for querying FEP contract data
Enhanced bridge syncer to support querying claims by global index and exit roots by hash
Moved EmptyLER constant to a shared location and updated references

Reviewed Changes

Copilot reviewed 24 out of 24 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
aggsender/query/certificate_query.go	Implements the core CertificateQuerier logic to calculate last settled blocks from multiple sources
aggsender/query/aggchain_fep_rollup_query.go	Provides FEP contract querying capabilities with no-op fallback for PP networks
bridgesync/processor.go	Adds GetClaimByGlobalIndex method for retrieving claims by their global index
aggsender/types/interfaces.go	Defines new interfaces for certificate querying and FEP contract interaction
agglayer/types/types.go	Adds ToBigInt method to GlobalIndex for converting to big.Int representation

bridgesync/processor.go

aggsender/query/certificate_query.go

agglayer/types/types.go

agglayer/mock_agglayer_client.go

bridgesync/processor.go

bridgesync/bridgesync_test.go

bridgesync/bridgesync.go

aggsender/query/certificate_query.go

sonarqubecloud · 2025-08-08T11:18:06Z

Quality Gate passed

Issues
3 New issues
0 Accepted issues

Measures
0 Security Hotspots
85.1% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

## 🔄 Changes Summary This PR decouples `aggsender-validator` from using the `Metadata` field in certificate to figure out block range. The validator will now use the `certificate querier` interface to get the `ToBlock` from the last settled certificate in order to figure out if the new certificate is valid or not. ## ⚠️ Breaking Changes NA ## 📋 Config Updates NA ## ✅ Testing - 🤖 **Automatic**: `aggkit` CI ## 🔗 Related PRs - #838 - #839

## 🔄 Changes Summary This PR completely removes the logic around setting `Metadata` field on certificates since it will no longer be used to determin block ranges on `aggsender`. `Metadata` field will still exist on certificates, but it will no longer be used, and a `ZeroHash` will be enforced instead. ## ⚠️ Breaking Changes NA ## 📋 Config Updates NA ## ✅ Testing - 🤖 **Automatic**: `aggkit` CI ## 🐞 Issues #799 ## 🔗 Related PRs #838 #839 #846

## 🔄 Changes Summary Enable the AggSender to work with multiple validator nodes in a committee-based validation system. The implementation includes validator services, multisig committee management, certificate validation improvements, and enhanced gRPC communication protocols. **Multisig Committee Support:** - Added `MultisigCommittee` type to manage signer sets and enforce signature thresholds. - Signers are represented as `SignerInfo` structs with both address and URL for improved context and error reporting. - Committee construction validates non-empty membership and non-zero threshold, preventing misconfiguration. - Dynamic signer management: methods for adding signers, duplicate checks by address and URL. **Aggsender Validator Refactor:** - The Aggsender certificate validation flow was refactored to integrate multisig logic. - The multisig validation logic is applicable to both `PP` and `FEP` certificates. - Certificate validation now checks for contiguous certificates, last L2 block, and settlement status using new queries. - Import bridge exit proof verification is handled via new logic using `verifyClaimProofs`, ensuring only valid proofs pass. **Certificate Metadata removal:** - It is gone from the agglayer and therefore it is not sent anymore from the aggsender either. - Only thing worth noting is that, when calculating `CertificateID`, instead of metadata field, which was used previously, we now use `ZeroHash`. **Smart contracts integration:** - **AggchainFEP contract:** Removed querying of `TrustedSequencer` address and rely on the signers committee instead - **AggchainBase contract:** Retrieve the multisig committee from the `AggchainBase` contract **Agglayer integration:** - Invoke the `GetNetworkState` API from agglayer to get the latest settled imported bridge exit info. - Multisig is populated into the certificate and sent to the Agglayer's `SendCertificate` gRPC endpoint ## ⚠️ Breaking Changes - 🛠️ **Config**: Make sure that `Mode` on the `Validator` and `AggSender` are the same. - 🔌 **API/CLI**: `aggkit` version (`v0.7.0`) that supports `multisig` will now require updated contracts to run. At least version `v12.1.0-rc.3` of `agglayer-contracts`, and a new version of `agglayer` which supports `multisig`, which is the `v0.4.0` of `agglayer`. - 🗑️ **Deprecated Features**: Aggsender Phase II validator signing logic ## 📋 Config Updates - Added `AggSender.RequireCommitteeMembershipCheck = false` parameter, which defines if a check on `aggsender proposer` startup will be performed to see if the proposer is in the `multisig` committee. - Added `Validator.RequireCommitteeMembershipCheck = {{AggSender.RequireCommitteeMembershipCheck}}` parameter, which defines if a check on `aggsender validator` startup will be performed to see if the validator is in the `multisig` committee. - Added `Validator.Mode = "PessimisticProof"` parameter, which acts the same as the `AggSender.Mode`. It tells the validator that the network is a `PP` network or an `FEP` network. It has to be the same as on `aggsender proposer`. - Added `Validator.FEPConfig.SovereignRollupAddr = "{{AggSender.SovereignRollupAddr}}" parameter which is the address of the `AggchainFEP` rollup on L1 for given network for which validator is running. - Added `Validator.FEPConfig.RequireNoBlockGap = {{AggSender.RequireNoFEPBlockGap}}, which acts the same as the given paremeter on `AggSender` (proposer) config, and tells the validator if gaps in blocks in certificates are allowed in `FEP` network. ```toml [AggSender] RequireCommitteeMembershipCheck = false [Validator] # PessimisticProof or AggchainProof Mode = "PessimisticProof" RequireCommitteeMembershipCheck = {{AggSender.RequireCommitteeMembershipCheck}} [Validator.FEPConfig] SovereignRollupAddr = "{{AggSender.SovereignRollupAddr}}" RequireNoBlockGap = "{{AggSender.RequireNoFEPBlockGap}}" ``` ## ✅ Testing - 🤖 **Automatic**: `aggkit` CI - 🖱️ **Manual**: [Optional: Steps to verify] ## 🐞 Issues - Closes #792 ## 🔗 Related PRs - #814 - #832 - #838 - #839 - #843 - #842 - #846 - #858 - #847 - #865 - #861 - #863 - #875 - #876 - #881 - #877 - #898 - #920 - #913 - #926 - #945 - #951 - #954 - #957 - #955 - #974 - #978 - #985 - #989 - #984 - #998 - #1017 - #1028 - #1034 - #1024 - #1052 - #1067 - #1068 - #1050 - #1071 - #1072 - #1060 - #1087 - #1077 - #1073 --------- Co-authored-by: Goran Rojovic <[email protected]> Co-authored-by: Goran Rojovic <[email protected]> Co-authored-by: Joan Esteban <[email protected]> Co-authored-by: Rachit Sonthalia <[email protected]> Co-authored-by: Arpit Temani <[email protected]> Co-authored-by: Copilot <[email protected]>

goran-ethernal self-assigned this Aug 7, 2025

goran-ethernal added the aggsender label Aug 7, 2025

goran-ethernal marked this pull request as ready for review August 7, 2025 10:02

goran-ethernal requested a review from Stefan-Ethernal August 7, 2025 10:03

goran-ethernal force-pushed the feat/compute-block-range-by-not-using-metadata branch from 7ae3acd to 1a71b93 Compare August 7, 2025 10:13

goran-ethernal added this to the AggKit - v0.8.0 milestone Aug 7, 2025

goran-ethernal requested a review from Copilot August 7, 2025 12:48

Copilot AI reviewed Aug 7, 2025

View reviewed changes

bridgesync/processor.go Show resolved Hide resolved

aggsender/query/certificate_query.go Show resolved Hide resolved

goran-ethernal added aggsender-proposer aggsender-validator labels Aug 7, 2025

Stefan-Ethernal approved these changes Aug 8, 2025

View reviewed changes

goran-ethernal added 3 commits August 8, 2025 13:08

feat: block range and aggchain fep querier

edca85d

fix: copilot comments

9f80148

fix: comments

acf7062

goran-ethernal force-pushed the feat/compute-block-range-by-not-using-metadata branch from a0a491d to acf7062 Compare August 8, 2025 11:08

goran-ethernal merged commit ff90cf9 into feat/aggsender-multisig Aug 8, 2025
21 of 22 checks passed

goran-ethernal deleted the feat/compute-block-range-by-not-using-metadata branch August 8, 2025 11:46

goran-ethernal mentioned this pull request Aug 11, 2025

feat: decouple aggsender-validator from metadata field #846

Merged

goran-ethernal mentioned this pull request Aug 12, 2025

feat: remove metadata from aggsender #858

Merged

goran-ethernal mentioned this pull request Aug 12, 2025

feat: adapt field from_block from code to infer from previousCert #799

Closed

Stefan-Ethernal mentioned this pull request Sep 9, 2025

feat: Aggsender multisig #860

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

feat: mechanism to calculate last settled block by not using `metadata` #838

feat: mechanism to calculate last settled block by not using `metadata` #838

Uh oh!

goran-ethernal commented Aug 7, 2025 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

sonarqubecloud bot commented Aug 8, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

feat: mechanism to calculate last settled block by not using metadata #838

feat: mechanism to calculate last settled block by not using metadata #838

Uh oh!

Conversation

goran-ethernal commented Aug 7, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🔄 Changes Summary

How the last settled block is calculated

⚠️ Breaking Changes

📋 Config Updates

✅ Testing

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Reviewed Changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

sonarqubecloud bot commented Aug 8, 2025

Quality Gate passed

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

feat: mechanism to calculate last settled block by not using `metadata` #838

feat: mechanism to calculate last settled block by not using `metadata` #838

goran-ethernal commented Aug 7, 2025 •

edited

Loading