Initsync skip local blobs (#13827)

* wip - init-sync skip available blob req

* satisfy deep source

* gaz

* don't need to sort blobs; simplify blobRequest stack

* wip debug log to watch blob skip behavior

* unit tests for new blob req generator

* refactor to reduce blob req func count

* log when WaitForSummarizer fails

---------

Co-authored-by: Kasey Kirkham <kasey@users.noreply.github.com>

beacon-chain/das/availability.go

@@ -94,6 +94,15 @@ func (s *LazilyPersistentStore) IsDataAvailable(ctx context.Context, current pri
 	entry := s.cache.ensure(key)
 	defer s.cache.delete(key)
 	root := b.Root()
+	sumz, err := s.store.WaitForSummarizer(ctx)
+	if err != nil {
+		log.WithField("root", fmt.Sprintf("%#x", b.Root())).
+			WithError(err).
+			Debug("Failed to receive BlobStorageSummarizer within IsDataAvailable")
+	} else {
+		entry.setDiskSummary(sumz.Summary(root))
+	}
+
 	// Verify we have all the expected sidecars, and fail fast if any are missing or inconsistent.
 	// We don't try to salvage problematic batches because this indicates a misbehaving peer and we'd rather
 	// ignore their response and decrease their peer score.

beacon-chain/das/cache.go

@@ -4,6 +4,7 @@ import (
 	"bytes"
 
 	"github.com/pkg/errors"
+	"github.com/prysmaticlabs/prysm/v5/beacon-chain/db/filesystem"
 	fieldparams "github.com/prysmaticlabs/prysm/v5/config/fieldparams"
 	"github.com/prysmaticlabs/prysm/v5/consensus-types/blocks"
 	"github.com/prysmaticlabs/prysm/v5/consensus-types/primitives"
@@ -59,7 +60,12 @@ func (c *cache) delete(key cacheKey) {
 
 // cacheEntry holds a fixed-length cache of BlobSidecars.
 type cacheEntry struct {
-	scs [fieldparams.MaxBlobsPerBlock]*blocks.ROBlob
+	scs         [fieldparams.MaxBlobsPerBlock]*blocks.ROBlob
+	diskSummary filesystem.BlobStorageSummary
+}
+
+func (e *cacheEntry) setDiskSummary(sum filesystem.BlobStorageSummary) {
+	e.diskSummary = sum
 }
 
 // stash adds an item to the in-memory cache of BlobSidecars.
@@ -81,9 +87,17 @@ func (e *cacheEntry) stash(sc *blocks.ROBlob) error {
 // the cache do not match those found in the block. If err is nil, then all expected
 // commitments were found in the cache and the sidecar slice return value can be used
 // to perform a DA check against the cached sidecars.
+// filter only returns blobs that need to be checked. Blobs already available on disk will be excluded.
 func (e *cacheEntry) filter(root [32]byte, kc safeCommitmentArray) ([]blocks.ROBlob, error) {
+	if e.diskSummary.AllAvailable(kc.count()) {
+		return nil, nil
+	}
 	scs := make([]blocks.ROBlob, kc.count())
 	for i := uint64(0); i < fieldparams.MaxBlobsPerBlock; i++ {
+		// We already have this blob, we don't need to write it or validate it.
+		if e.diskSummary.HasIndex(i) {
+			continue
+		}
 		if kc[i] == nil {
 			if e.scs[i] != nil {
 				return nil, errors.Wrapf(errCommitmentMismatch, "root=%#x, index=%#x, commitment=%#x, no block commitment", root, i, e.scs[i].KzgCommitment)