batcher: add batchSubmitter.checkExpectedProgress (ethereum-optimism#…

…12430)

* implement batchSubmitter.checkExpectedProgress

* remove buffer variable

* add warning logs when calling waitNodeSyncAndClearState

* push method down into channel manager and add test

* clarify SyncStatus documentation

* improve TestChannelManager_CheckExpectedProgress

make parameters "tighter" / more realistic and check an extra case

op-batcher/batcher/channel_manager.go

@@ -549,3 +549,19 @@ func (s *channelManager) PendingDABytes() int64 {
 	}
 	return int64(f)
 }
+
+// CheckExpectedProgress uses the supplied syncStatus to infer
+// whether the node providing the status has made the expected
+// safe head progress given fully submitted channels held in
+// state.
+func (m *channelManager) CheckExpectedProgress(syncStatus eth.SyncStatus) error {
+	for _, ch := range m.channelQueue {
+		if ch.isFullySubmitted() && // This implies a number of l1 confirmations has passed, depending on how the txmgr was configured
+			!ch.isTimedOut() &&
+			syncStatus.CurrentL1.Number > ch.maxInclusionBlock &&
+			syncStatus.SafeL2.Number < ch.LatestL2().Number {
+			return errors.New("safe head did not make expected progress")
+		}
+	}
+	return nil
+}

op-batcher/batcher/channel_manager_test.go

@@ -627,3 +627,57 @@ func TestChannelManager_ChannelOutFactory(t *testing.T) {
 
 	require.IsType(t, &ChannelOutWrapper{}, m.currentChannel.channelBuilder.co)
 }
+
+func TestChannelManager_CheckExpectedProgress(t *testing.T) {
+	l := testlog.Logger(t, log.LevelCrit)
+	cfg := channelManagerTestConfig(100, derive.SingularBatchType)
+	cfg.InitNoneCompressor()
+	m := NewChannelManager(l, metrics.NoopMetrics, cfg, defaultTestRollupConfig)
+
+	channelMaxInclusionBlockNumber := uint64(3)
+	channelLatestSafeBlockNumber := uint64(11)
+
+	// Prepare a (dummy) fully submitted channel
+	// with
+	// maxInclusionBlock and latest safe block number as above
+	A, err := newChannelWithChannelOut(l, metrics.NoopMetrics, cfg, m.rollupCfg, 0)
+	require.NoError(t, err)
+	rng := rand.New(rand.NewSource(123))
+	a0 := derivetest.RandomL2BlockWithChainId(rng, 1, defaultTestRollupConfig.L2ChainID)
+	a0 = a0.WithSeal(&types.Header{Number: big.NewInt(int64(channelLatestSafeBlockNumber))})
+	_, err = A.AddBlock(a0)
+	require.NoError(t, err)
+	A.maxInclusionBlock = channelMaxInclusionBlockNumber
+	A.Close()
+	A.channelBuilder.frames = nil
+	A.channelBuilder.frameCursor = 0
+	require.True(t, A.isFullySubmitted())
+
+	m.channelQueue = append(m.channelQueue, A)
+
+	// The current L1 number implies that
+	// channel A above should have been derived
+	// from, so we expect safe head to progress to
+	// the channelLatestSafeBlockNumber.
+	// Since the safe head moved to 11, there is no error:
+	ss := eth.SyncStatus{
+		CurrentL1: eth.L1BlockRef{Number: channelMaxInclusionBlockNumber + 1},
+		SafeL2:    eth.L2BlockRef{Number: channelLatestSafeBlockNumber},
+	}
+	err = m.CheckExpectedProgress(ss)
+	require.NoError(t, err)
+
+	// If the currentL1 is as above but the
+	// safe head is less than channelLatestSafeBlockNumber,
+	// the method should return an error:
+	ss.SafeL2 = eth.L2BlockRef{Number: channelLatestSafeBlockNumber - 1}
+	err = m.CheckExpectedProgress(ss)
+	require.Error(t, err)
+
+	// If the safe head is still less than channelLatestSafeBlockNumber
+	// but the currentL1 is _equal_ to the channelMaxInclusionBlockNumber
+	// there should be no error as that block is still being derived from:
+	ss.CurrentL1 = eth.L1BlockRef{Number: channelMaxInclusionBlockNumber}
+	err = m.CheckExpectedProgress(ss)
+	require.NoError(t, err)
+}

op-batcher/batcher/driver.go

@@ -471,17 +471,20 @@ func (l *BatchSubmitter) mainLoop(ctx context.Context, receiptsCh chan txmgr.TxR
 
 			l.state.pruneSafeBlocks(syncStatus.SafeL2)
 			l.state.pruneChannels(syncStatus.SafeL2)
+
+			err = l.state.CheckExpectedProgress(*syncStatus)
+			if err != nil {
+				l.Log.Warn("error checking expected progress, clearing state and waiting for node sync", "err", err)
+				l.waitNodeSyncAndClearState()
+				continue
+			}
+
 			if err := l.loadBlocksIntoState(*syncStatus, l.shutdownCtx); errors.Is(err, ErrReorg) {
-				// Wait for any in flight transactions
-				// to be ingested by the node before
-				// we start loading blocks again.
-				err := l.waitNodeSync()
-				if err != nil {
-					l.Log.Warn("error waiting for node sync", "err", err)
-				}
-				l.clearState(l.shutdownCtx)
+				l.Log.Warn("error loading blocks, clearing state and waiting for node sync", "err", err)
+				l.waitNodeSyncAndClearState()
 				continue
 			}
+
 			l.publishStateToL1(queue, receiptsCh, daGroup, l.Config.PollInterval)
 		case <-ctx.Done():
 			l.Log.Warn("main loop returning")
@@ -579,6 +582,17 @@ func (l *BatchSubmitter) throttlingLoop(ctx context.Context) {
 	}
 }
 
+func (l *BatchSubmitter) waitNodeSyncAndClearState() {
+	// Wait for any in flight transactions
+	// to be ingested by the node before
+	// we start loading blocks again.
+	err := l.waitNodeSync()
+	if err != nil {
+		l.Log.Warn("error waiting for node sync", "err", err)
+	}
+	l.clearState(l.shutdownCtx)
+}
+
 // waitNodeSync Check to see if there was a batcher tx sent recently that
 // still needs more block confirmations before being considered finalized
 func (l *BatchSubmitter) waitNodeSync() error {

op-service/eth/sync_status.go

@@ -5,7 +5,7 @@ package eth
 type SyncStatus struct {
 	// CurrentL1 is the L1 block that the derivation process is last idled at.
 	// This may not be fully derived into L2 data yet.
-	// The safe L2 blocks were produced/included fully from the L1 chain up to and including this L1 block.
+	// The safe L2 blocks were produced/included fully from the L1 chain up to _but excluding_ this L1 block.
 	// If the node is synced, this matches the HeadL1, minus the verifier confirmation distance.
 	CurrentL1 L1BlockRef `json:"current_l1"`
 	// CurrentL1Finalized is a legacy sync-status attribute. This is deprecated.