feat(archiver): decouple calldata from blob fetching in L1 synchronizer by spalladino · Pull Request #22472 · AztecProtocol/aztec-packages

spalladino · 2026-04-10T17:36:08Z

Motivation

When the node is a validator that already built a checkpoint locally (via addProposedBlock + setProposedCheckpoint), the blocks are already in the archiver store. Fetching blobs from the beacon chain is redundant and expensive, especially during sync. This decouples calldata and blob retrieval so we can skip blob fetching when the proposed checkpoint matches.

Approach

Split retrieveCheckpointsFromRollup into two phases: (1) fetch calldata only (header, attestations, archive root), (2) check the archiver store for a proposed checkpoint with matching header. If found, promote it to confirmed via a fast path. Otherwise, fetch blobs in parallel (same asyncPool(10, ...) concurrency as before) and store as normal.

Changes

archiver (l1/data_retrieval.ts): New CalldataOnlyCheckpoint type, retrieveCheckpointCalldataFromRollup (calldata-only fetch), and fetchBlobsAndBuildPublishedCheckpoint (deferred blob fetch). Existing functions left intact.
archiver (store/block_store.ts): New promoteProposedToCheckpointed method that reads existing blocks from store, writes a confirmed checkpoint entry with L1 metadata + attestations, and clears the proposed singleton.
archiver (store/kv_archiver_store.ts): Pass-through for promoteProposedToCheckpointed.
archiver (modules/data_store_updater.ts): New promoteProposedCheckpoint wrapper that handles validation status and tips cache refresh.
archiver (modules/l1_synchronizer.ts): handleCheckpoints now partitions calldata checkpoints into promote-vs-fetch-blobs, fetches blobs in parallel for non-matching ones, promotes the matched one, then merges results for validation.

Fixes A-877

spalladino · 2026-04-10T20:17:36Z

Got to consider race condition where a proposer pushes their proposed checkpoint to its archiver before the previous checkpoint is checkpointed on L1, but this requires a larger change on the block store.

Maddiaa0 · 2026-04-21T10:30:43Z

+        ? undefined
+        : await this.store.getProposedCheckpointOnly();
+      const last = calldataCheckpoints[calldataCheckpoints.length - 1];
+      const toPromote =


to Promote is only used as a boolean? The logs can just use last directly?

Maddiaa0 · 2026-04-21T10:47:13Z

    description: 'Skip validating checkpoint attestations (for testing purposes only)',
    ...booleanConfigHelper(false),
  },
+  skipPromoteProposedCheckpointDuringL1Sync: {


is it worth prefixing this option with test_

Maddiaa0 · 2026-04-21T11:22:00Z

+        }
+
+        // And persist the promoted checkpoint if it passed validation
+        if (toPromote && validCheckpoints.some(c => c.checkpoint.number === toPromote.checkpointNumber)) {


this .some and .find below are doing the same search. could probably extract it

Maddiaa0 · 2026-04-21T11:27:33Z

        const [processDuration, result] = await elapsed(() =>
          execInSpan(this.tracer, 'Archiver.addCheckpoints', () =>
-            this.updater.addCheckpoints(validCheckpoints, updatedValidationResult),
+            this.updater.addCheckpoints(checkpointsToAdd, updatedValidationResult),


should addCheckpoints + promoteProposed be able to be updated within the same call? Ideally these should update atomically? could have some kind of queue + flush

Maddiaa0 · 2026-04-21T11:29:24Z

+        this.log.debug(
+          `Building published checkpoint from proposed ${toPromote.checkpointNumber} (skipping blob fetch)`,
+        );
+        const blocks = await this.store.getBlocks(BlockNumber(proposed.startBlock), proposed.blockCount);


just realised getBlocks does not assert that blockCount blocks are returned

Maddiaa0 · 2026-04-21T11:30:59Z

+    return await this.db.transactionAsync(async () => {
+      const proposed = await this.getProposedCheckpointOnly();
+      if (!proposed) {
+        throw new Error('Cannot promote proposed checkpoint: no proposed checkpoint exists');


not for now but this file has a big variance in custom error / text error

Maddiaa0

my comments are mainly nits

Splits checkpoint retrieval into two phases: first fetch calldata only (header, attestations, archive root), then check if a proposed checkpoint with matching header already exists in the store. If so, skip blob fetching and promote the proposed checkpoint to confirmed. Otherwise, fetch blobs in parallel as before. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

- Move promote to after attestation validation to avoid persisting invalid checkpoints (split into build + persist steps) - Add validateCheckpoint call in promote path - Add archive root comparison to proposed checkpoint match condition - Remove dead retrieveCheckpointsFromRollup and processCheckpointProposedLogs - Pass pendingChainValidationStatus to promote path Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

…er (#22716) Reopening #22472 which was accidentally merged ## Motivation When the node is a validator that already built a checkpoint locally (via `addProposedBlock` + `setProposedCheckpoint`), the blocks are already in the archiver store. Fetching blobs from the beacon chain is redundant and expensive, especially during sync. This decouples calldata and blob retrieval so we can skip blob fetching when the proposed checkpoint matches. ## Approach Split `retrieveCheckpointsFromRollup` into two phases: (1) fetch calldata only (header, attestations, archive root), (2) check the archiver store for a proposed checkpoint with matching header. If found, promote it to confirmed via a fast path. Otherwise, fetch blobs in parallel (same `asyncPool(10, ...)` concurrency as before) and store as normal. ## Changes - **archiver (l1/data_retrieval.ts)**: New `CalldataOnlyCheckpoint` type, `retrieveCheckpointCalldataFromRollup` (calldata-only fetch), and `fetchBlobsAndBuildPublishedCheckpoint` (deferred blob fetch). Existing functions left intact. - **archiver (store/block_store.ts)**: New `promoteProposedToCheckpointed` method that reads existing blocks from store, writes a confirmed checkpoint entry with L1 metadata + attestations, and clears the proposed singleton. - **archiver (store/kv_archiver_store.ts)**: Pass-through for `promoteProposedToCheckpointed`. - **archiver (modules/data_store_updater.ts)**: New `promoteProposedCheckpoint` wrapper that handles validation status and tips cache refresh. - **archiver (modules/l1_synchronizer.ts)**: `handleCheckpoints` now partitions calldata checkpoints into promote-vs-fetch-blobs, fetches blobs in parallel for non-matching ones, promotes the matched one, then merges results for validation. Fixes A-877 --------- Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

spalladino added ci-no-fail-fast Sets NO_FAIL_FAST in the CI so the run is not aborted on the first failure ci-draft Run CI on draft PRs. labels Apr 10, 2026

spalladino marked this pull request as ready for review April 13, 2026 21:33

spalladino removed the ci-draft Run CI on draft PRs. label Apr 13, 2026

spalladino requested a review from Maddiaa0 April 20, 2026 20:40

Maddiaa0 reviewed Apr 21, 2026

View reviewed changes

Maddiaa0 approved these changes Apr 21, 2026

View reviewed changes

spalladino force-pushed the spl/decouple-calldata-blobs branch 2 times, most recently from 6e9e318 to 52250fe Compare April 21, 2026 18:34

spalladino enabled auto-merge (squash) April 21, 2026 18:34

spalladino and others added 6 commits April 21, 2026 18:16

refactor and e2e test

d35b8c2

small fixes

c8911e3

fix unit test

3873aaa

address code review changes

20c5f7c

spalladino force-pushed the spl/decouple-calldata-blobs branch from 52250fe to 20c5f7c Compare April 21, 2026 21:16

spalladino changed the base branch from merge-train/spartan to spl/fix-genesis-world-state-again April 21, 2026 21:16

spalladino merged commit 48236ce into spl/fix-genesis-world-state-again Apr 21, 2026
9 of 10 checks passed

spalladino deleted the spl/decouple-calldata-blobs branch April 21, 2026 21:16

spalladino mentioned this pull request Apr 21, 2026

Revert "feat(archiver): decouple calldata from blob fetching in L1 synchronizer" #22715

Closed

spalladino restored the spl/decouple-calldata-blobs branch April 21, 2026 21:18

spalladino mentioned this pull request Apr 21, 2026

feat(archiver): decouple calldata from blob fetching in L1 synchronizer #22716

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(archiver): decouple calldata from blob fetching in L1 synchronizer#22472

feat(archiver): decouple calldata from blob fetching in L1 synchronizer#22472
spalladino merged 6 commits into
spl/fix-genesis-world-state-againfrom
spl/decouple-calldata-blobs

spalladino commented Apr 10, 2026

Uh oh!

spalladino commented Apr 10, 2026 •

edited

Loading

Uh oh!

Maddiaa0 Apr 21, 2026

Uh oh!

Maddiaa0 Apr 21, 2026

Uh oh!

Maddiaa0 Apr 21, 2026

Uh oh!

Maddiaa0 Apr 21, 2026

Uh oh!

Maddiaa0 Apr 21, 2026

Uh oh!

Maddiaa0 Apr 21, 2026

Uh oh!

Maddiaa0 left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

spalladino commented Apr 10, 2026

Motivation

Approach

Changes

Uh oh!

spalladino commented Apr 10, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Maddiaa0 Apr 21, 2026

Choose a reason for hiding this comment

Uh oh!

Maddiaa0 Apr 21, 2026

Choose a reason for hiding this comment

Uh oh!

Maddiaa0 Apr 21, 2026

Choose a reason for hiding this comment

Uh oh!

Maddiaa0 Apr 21, 2026

Choose a reason for hiding this comment

Uh oh!

Maddiaa0 Apr 21, 2026

Choose a reason for hiding this comment

Uh oh!

Maddiaa0 Apr 21, 2026

Choose a reason for hiding this comment

Uh oh!

Maddiaa0 left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

spalladino commented Apr 10, 2026 •

edited

Loading