PeerDAS: Implement syncing in a disjoint network (Also know as "perfect PeerDAS" network). (#15644)

* `computeIndicesByRootByPeer`: Add 1 slack epoch regarding peer head slot.

* `FetchDataColumnSidecars`: Switch mode.

Before this commit, this function returned on error as long as at least ONE requested sidecar was not retrieved.

Now, this function retrieves what it can (best effort mode) and returns an additional value which is the map of missing sidecars after running this function.

It is now the role of the caller to check this extra returned value and decide what to do in case some requested sidecars are still missing.

* `fetchOriginDataColumnSidecars`: Optimize

Before this commit, when running `fetchOriginDataColumnSidecars`, all the missing sidecars had to been retrieved in a single shot for the sidecars to be considered as available. The issue was, if for example `sync.FetchDataColumnSidecars` returned all but one sidecar, the returned sidecars were NOT saved, and on the next iteration, all the previously fetched sidecars had to be requested again (from peers.)

After this commit, we greedily save all fetched sidecars, solving this issue.

* Initial sync: Do not fetch data column sidecars before the retention period.

* Implement perfect peerdas syncing.

* Add changelog.

* Fix James' comment.

* Fix James' comment.

* Fix James' comment.

* Fix James' comment.

* Fix James' comment.

* Fix James' comment.

* Fix James' comment.

* Update beacon-chain/sync/data_column_sidecars.go

Co-authored-by: Potuz <potuz@prysmaticlabs.com>

* Update beacon-chain/sync/data_column_sidecars.go

Co-authored-by: Potuz <potuz@prysmaticlabs.com>

* Update beacon-chain/sync/data_column_sidecars.go

Co-authored-by: Potuz <potuz@prysmaticlabs.com>

* Update after Potuz's comment.

* Fix Potuz's commit.

* Fix James' comment.

---------

Co-authored-by: Potuz <potuz@prysmaticlabs.com>

beacon-chain/sync/initial-sync/service.go

@@ -22,7 +22,6 @@ import (
 	"github.com/OffchainLabs/prysm/v6/beacon-chain/sync"
 	"github.com/OffchainLabs/prysm/v6/beacon-chain/verification"
 	"github.com/OffchainLabs/prysm/v6/cmd/beacon-chain/flags"
-	fieldparams "github.com/OffchainLabs/prysm/v6/config/fieldparams"
 	"github.com/OffchainLabs/prysm/v6/config/params"
 	"github.com/OffchainLabs/prysm/v6/consensus-types/blocks"
 	"github.com/OffchainLabs/prysm/v6/crypto/rand"
@@ -237,14 +236,14 @@ func (s *Service) fetchOriginSidecars(peers []peer.ID) error {
 	blockVersion := roBlock.Version()
 
 	if blockVersion >= version.Fulu {
-		if err := s.fetchOriginColumns(roBlock, delay); err != nil {
+		if err := s.fetchOriginDataColumnSidecars(roBlock, delay); err != nil {
 			return errors.Wrap(err, "fetch origin columns")
 		}
 		return nil
 	}
 
 	if blockVersion >= version.Deneb {
-		if err := s.fetchOriginBlobs(peers, roBlock); err != nil {
+		if err := s.fetchOriginBlobSidecars(peers, roBlock); err != nil {
 			return errors.Wrap(err, "fetch origin blobs")
 		}
 	}
@@ -356,7 +355,7 @@ func missingBlobRequest(blk blocks.ROBlock, store *filesystem.BlobStorage) (p2pt
 	return req, nil
 }
 
-func (s *Service) fetchOriginBlobs(pids []peer.ID, rob blocks.ROBlock) error {
+func (s *Service) fetchOriginBlobSidecars(pids []peer.ID, rob blocks.ROBlock) error {
 	r := rob.Root()
 
 	req, err := missingBlobRequest(rob, s.cfg.BlobStorage)
@@ -394,11 +393,12 @@ func (s *Service) fetchOriginBlobs(pids []peer.ID, rob blocks.ROBlock) error {
 	return fmt.Errorf("no connected peer able to provide blobs for checkpoint sync block %#x", r)
 }
 
-func (s *Service) fetchOriginColumns(roBlock blocks.ROBlock, delay time.Duration) error {
+func (s *Service) fetchOriginDataColumnSidecars(roBlock blocks.ROBlock, delay time.Duration) error {
 	const (
 		errorMessage     = "Failed to fetch origin data column sidecars"
 		warningIteration = 10
 	)
+
 	samplesPerSlot := params.BeaconConfig().SamplesPerSlot
 
 	// Return early if the origin block has no blob commitments.
@@ -411,7 +411,7 @@ func (s *Service) fetchOriginColumns(roBlock blocks.ROBlock, delay time.Duration
 		return nil
 	}
 
-	// Compute the columns to request.
+	// Compute the indices we need to custody.
 	custodyGroupCount, err := s.cfg.P2P.CustodyGroupCount()
 	if err != nil {
 		return errors.Wrap(err, "custody group count")
@@ -423,9 +423,30 @@ func (s *Service) fetchOriginColumns(roBlock blocks.ROBlock, delay time.Duration
 		return errors.Wrap(err, "fetch peer info")
 	}
 
-	// Fetch origin data column sidecars.
 	root := roBlock.Root()
 
+	log := log.WithFields(logrus.Fields{
+		"blockRoot":       fmt.Sprintf("%#x", roBlock.Root()),
+		"blobCount":       len(commitments),
+		"dataColumnCount": len(info.CustodyColumns),
+	})
+
+	// Check if some needed data column sidecars are missing.
+	stored := s.cfg.DataColumnStorage.Summary(root).Stored()
+	missing := make(map[uint64]bool, len(info.CustodyColumns))
+	for column := range info.CustodyColumns {
+		if !stored[column] {
+			missing[column] = true
+		}
+	}
+
+	if len(missing) == 0 {
+		// All needed data column sidecars are present, exit early.
+		log.Info("All needed origin data column sidecars are already present")
+
+		return nil
+	}
+
 	params := sync.DataColumnSidecarsParams{
 		Ctx:                     s.ctx,
 		Tor:                     s.clock,
@@ -436,46 +457,39 @@ func (s *Service) fetchOriginColumns(roBlock blocks.ROBlock, delay time.Duration
 		DownscorePeerOnRPCFault: true,
 	}
 
-	var verifiedRoDataColumnsByRoot map[[fieldparams.RootLength]byte][]blocks.VerifiedRODataColumn
 	for attempt := uint64(0); ; attempt++ {
-		verifiedRoDataColumnsByRoot, err = sync.FetchDataColumnSidecars(params, []blocks.ROBlock{roBlock}, info.CustodyColumns)
-		if err == nil {
-			break
+		// Retrieve missing data column sidecars.
+		verifiedRoSidecarsByRoot, missingIndicesByRoot, err := sync.FetchDataColumnSidecars(params, []blocks.ROBlock{roBlock}, missing)
+		if err != nil {
+			return errors.Wrap(err, "fetch data column sidecars")
 		}
 
-		log := log.WithError(err).WithFields(logrus.Fields{
-			"attempt": attempt,
-			"delay":   delay,
+		// Save retrieved data column sidecars.
+		if err := s.cfg.DataColumnStorage.Save(verifiedRoSidecarsByRoot[root]); err != nil {
+			return errors.Wrap(err, "save data column sidecars")
+		}
+
+		// Check if some needed data column sidecars are missing.
+		if len(missingIndicesByRoot) == 0 {
+			log.Info("Retrieved all needed origin data column sidecars")
+
+			return nil
+		}
+
+		// Some sidecars are still missing.
+		log := log.WithFields(logrus.Fields{
+			"attempt":        attempt,
+			"missingIndices": sortedSliceFromMap(missingIndicesByRoot[root]),
+			"delay":          delay,
 		})
 
-		if attempt%warningIteration == 0 && attempt > 0 {
-			log.Warning(errorMessage)
-			time.Sleep(delay)
-
-			continue
+		logFunc := log.Debug
+		if attempt > 0 && attempt%warningIteration == 0 {
+			logFunc = log.Warning
 		}
 
-		log.Debug(errorMessage)
-		time.Sleep(delay)
+		logFunc("Failed to fetch some origin data column sidecars, retrying later")
 	}
-
-	// Save origin data columns to disk.
-	verifiedRoDataColumnsSidecars, ok := verifiedRoDataColumnsByRoot[root]
-	if !ok {
-		return fmt.Errorf("cannot extract origins data column sidecars for block root %#x - should never happen", root)
-	}
-
-	if err := s.cfg.DataColumnStorage.Save(verifiedRoDataColumnsSidecars); err != nil {
-		return errors.Wrap(err, "save data column sidecars")
-	}
-
-	log.WithFields(logrus.Fields{
-		"blockRoot":   fmt.Sprintf("%#x", roBlock.Root()),
-		"blobCount":   len(commitments),
-		"columnCount": len(verifiedRoDataColumnsSidecars),
-	}).Info("Successfully downloaded data column sidecars for checkpoint sync block")
-
-	return nil
 }
 
 func shufflePeers(pids []peer.ID) {