@@ -27,6 +27,7 @@ if err := a.verifyPoolL2TxWrite(writeTx); err != nil {
 		retBadReq(err, c)
 		return
 	}
+	writeTx.ClientIP = c.ClientIP()
 	// Insert to DB
 	if err := a.l2.AddTxAPI(writeTx); err != nil {
 		retSQLErr(err, c)

@@ -1,2 +1,3 @@
 cfg.example.secret.toml
 cfg.toml
+node

@@ -32,6 +32,7 @@
 [Synchronizer]
 SyncLoopInterval = "1s"
 StatsRefreshPeriod = "1s"
+StoreAccountUpdates = true
 
 [SmartContracts]
 Rollup   = "0x8EEaea23686c319133a7cC110b840d1591d9AeE0"
@@ -55,6 +56,7 @@ ForgeRetryInterval = "500ms"
 SyncRetryInterval = "1s"
 ForgeDelay = "10s"
 ForgeNoTxsDelay = "0s"
+PurgeByExtDelInterval = "1m"
 
 [Coordinator.FeeAccount]
 Address = "0x56232B1c5B10038125Bc7345664B4AFD745bcF8E"

@@ -91,6 +91,10 @@ ForgeNoTxsDelay Duration `validate:"-"`
 	// SyncRetryInterval is the waiting interval between calls to the main
 	// handler of a synced block after an error
 	SyncRetryInterval Duration `validate:"required"`
+	// PurgeByExtDelInterval is the waiting interval between calls
+	// to the PurgeByExternalDelete function of the l2db which deletes
+	// pending txs externally marked by the column `external_delete`
+	PurgeByExtDelInterval Duration `validate:"required"`
 	// L2DB is the DB that holds the pool of L2Txs
 	L2DB struct {
 		// SafetyPeriod is the number of batches after which

@@ -85,6 +85,10 @@ ForgeNoTxsDelay time.Duration
 	// SyncRetryInterval is the waiting interval between calls to the main
 	// handler of a synced block after an error
 	SyncRetryInterval time.Duration
+	// PurgeByExtDelInterval is the waiting interval between calls
+	// to the PurgeByExternalDelete function of the l2db which deletes
+	// pending txs externally marked by the column `external_delete`
+	PurgeByExtDelInterval time.Duration
 	// EthClientAttemptsDelay is delay between attempts do do an eth client
 	// RPC call
 	EthClientAttemptsDelay time.Duration
@@ -153,6 +157,15 @@ ctx    context.Context
 	wg     sync.WaitGroup
 	cancel context.CancelFunc
 
+	// mutexL2DBUpdateDelete protects updates to the L2DB so that
+	// these two processes always happen exclusively:
+	// - Pipeline taking pending txs, running through the TxProcessor and
+	//   marking selected txs as forging
+	// - Coordinator deleting pending txs that have been marked with
+	//   `external_delete`.
+	// Without this mutex, the coordinator could delete a pending txs that
+	// has just been selected by the TxProcessor in the pipeline.
+	mutexL2DBUpdateDelete sync.Mutex
 	pipeline              *Pipeline
 	lastNonFailedBatchNum common.BatchNum
 
@@ -248,7 +261,8 @@
 func (c *Coordinator) newPipeline(ctx context.Context) (*Pipeline, error) {
 	c.pipelineNum++
 	return NewPipeline(ctx, c.cfg, c.pipelineNum, c.historyDB, c.l2DB, c.txSelector,
-		c.batchBuilder, c.purger, c, c.txManager, c.provers, &c.consts)
+		c.batchBuilder, &c.mutexL2DBUpdateDelete, c.purger, c, c.txManager,
+		c.provers, &c.consts)
 }
 
 // MsgSyncBlock indicates an update to the Synchronizer stats
@@ -524,6 +538,24 @@ waitCh = time.After(c.cfg.SyncRetryInterval)
 					continue
 				}
 				waitCh = time.After(longWaitDuration)
+			}
+		}
+	}()
+
+	c.wg.Add(1)
+	go func() {
+		for {
+			select {
+			case <-c.ctx.Done():
+				log.Info("Coordinator L2DB.PurgeByExternalDelete loop done")
+				c.wg.Done()
+				return
+			case <-time.After(c.cfg.PurgeByExtDelInterval):
+				c.mutexL2DBUpdateDelete.Lock()
+				if err := c.l2DB.PurgeByExternalDelete(); err != nil {
+					log.Errorw("L2DB.PurgeByExternalDelete", "err", err)
+				}
+				c.mutexL2DBUpdateDelete.Unlock()
 			}
 		}
 	}()

@@ -45,15 +45,16 @@ rw            sync.RWMutex
 	errAtBatchNum common.BatchNum
 	lastForgeTime time.Time
 
-	proversPool  *ProversPool
-	provers      []prover.Client
-	coord        *Coordinator
-	txManager    *TxManager
-	historyDB    *historydb.HistoryDB
-	l2DB         *l2db.L2DB
-	txSelector   *txselector.TxSelector
-	batchBuilder *batchbuilder.BatchBuilder
-	purger       *Purger
+	proversPool           *ProversPool
+	provers               []prover.Client
+	coord                 *Coordinator
+	txManager             *TxManager
+	historyDB             *historydb.HistoryDB
+	l2DB                  *l2db.L2DB
+	txSelector            *txselector.TxSelector
+	batchBuilder          *batchbuilder.BatchBuilder
+	mutexL2DBUpdateDelete *sync.Mutex
+	purger                *Purger
 
 	stats       synchronizer.Stats
 	vars        synchronizer.SCVariables
@@ -84,6 +85,7 @@ historyDB *historydb.HistoryDB,
 	l2DB *l2db.L2DB,
 	txSelector *txselector.TxSelector,
 	batchBuilder *batchbuilder.BatchBuilder,
+	mutexL2DBUpdateDelete *sync.Mutex,
 	purger *Purger,
 	coord *Coordinator,
 	txManager *TxManager,
@@ -104,19 +106,20 @@ if proversPoolSize == 0 {
 		return nil, tracerr.Wrap(fmt.Errorf("no provers in the pool"))
 	}
 	return &Pipeline{
-		num:          num,
-		cfg:          cfg,
-		historyDB:    historyDB,
-		l2DB:         l2DB,
-		txSelector:   txSelector,
-		batchBuilder: batchBuilder,
-		provers:      provers,
-		proversPool:  proversPool,
-		purger:       purger,
-		coord:        coord,
-		txManager:    txManager,
-		consts:       *scConsts,
-		statsVarsCh:  make(chan statsVars, queueLen),
+		num:                   num,
+		cfg:                   cfg,
+		historyDB:             historyDB,
+		l2DB:                  l2DB,
+		txSelector:            txSelector,
+		batchBuilder:          batchBuilder,
+		provers:               provers,
+		proversPool:           proversPool,
+		mutexL2DBUpdateDelete: mutexL2DBUpdateDelete,
+		purger:                purger,
+		coord:                 coord,
+		txManager:             txManager,
+		consts:                *scConsts,
+		statsVarsCh:           make(chan statsVars, queueLen),
 	}, nil
 }
 
@@ -199,7 +202,9 @@ // handleForgeBatch calls p.forgeBatch to forge the batch and get the zkInputs,
 // and then waits for an available proof server and sends the zkInputs to it so
 // that the proof computation begins.
 func (p *Pipeline) handleForgeBatch(ctx context.Context, batchNum common.BatchNum) (*BatchInfo, error) {
+	p.mutexL2DBUpdateDelete.Lock()
 	batchInfo, err := p.forgeBatch(batchNum)
+	p.mutexL2DBUpdateDelete.Unlock()
 	if ctx.Err() != nil {
 		return nil, ctx.Err()
 	} else if err != nil {

@@ -73,24 +73,6 @@ addr,
 	))
 }
 
-// AddTx inserts a tx to the pool
-func (l2db *L2DB) AddTx(tx *PoolL2TxWrite) error {
-	row := l2db.db.QueryRow(
-		"SELECT COUNT(*) FROM tx_pool WHERE state = $1;",
-		common.PoolL2TxStatePending,
-	)
-	var totalTxs uint32
-	if err := row.Scan(&totalTxs); err != nil {
-		return tracerr.Wrap(err)
-	}
-	if totalTxs >= l2db.maxTxs {
-		return tracerr.New(
-			"The pool is at full capacity. More transactions are not accepted currently",
-		)
-	}
-	return tracerr.Wrap(meddler.Insert(l2db.db, "tx_pool", tx))
-}
-
 // UpdateTxsInfo updates the parameter Info of the pool transactions
 func (l2db *L2DB) UpdateTxsInfo(txs []common.PoolL2Tx) error {
 	if len(txs) == 0 {
@@ -354,3 +336,14 @@ time.Unix(now-int64(l2db.ttl.Seconds()), 0),
 	)
 	return tracerr.Wrap(err)
 }
+
+// PurgeByExternalDelete deletes all pending transactions marked with true in
+// the `external_delete` column.  An external process can set this column to
+// true to instruct the coordinator to delete the tx when possible.
+func (l2db *L2DB) PurgeByExternalDelete() error {
+	_, err := l2db.db.Exec(
+		`DELETE from tx_pool WHERE (external_delete = true AND state = $1);`,
+		common.PoolL2TxStatePending,
+	)
+	return tracerr.Wrap(err)
+}

@@ -1,6 +1,7 @@
 package l2db
 
 import (
+	"database/sql"
 	"math"
 	"math/big"
 	"os"
@@ -701,3 +702,56 @@ txs[i].AbsoluteFeeUpdate = dbTx.AbsoluteFeeUpdate
 		assert.Equal(t, txs[i], *dbTx)
 	}
 }
+
+func TestPurgeByExternalDelete(t *testing.T) {
+	err := prepareHistoryDB(historyDB)
+	if err != nil {
+		log.Error("Error prepare historyDB", err)
+	}
+	txs, err := generatePoolL2Txs()
+	assert.NoError(t, err)
+
+	// We will work with 8 txs
+	require.GreaterOrEqual(t, len(txs), 8)
+	txs = txs[:8]
+	for i := range txs {
+		require.NoError(t, l2DB.AddTxTest(&txs[i]))
+	}
+
+	// We will recreate this scenario:
+	// tx index, status , external_delete
+	// 0       , pending, false
+	// 1       , pending, false
+	// 2       , pending, true // will be deleted
+	// 3       , pending, true // will be deleted
+	// 4       , fging  , false
+	// 5       , fging  , false
+	// 6       , fging  , true
+	// 7       , fging  , true
+
+	require.NoError(t, l2DB.StartForging(
+		[]common.TxID{txs[4].TxID, txs[5].TxID, txs[6].TxID, txs[7].TxID},
+		1))
+	_, err = l2DB.db.Exec(
+		`UPDATE tx_pool SET external_delete = true WHERE
+			tx_id IN ($1, $2, $3, $4)
+		;`,
+		txs[2].TxID, txs[3].TxID, txs[6].TxID, txs[7].TxID,
+	)
+	require.NoError(t, err)
+	require.NoError(t, l2DB.PurgeByExternalDelete())
+
+	// Query txs that are have been not deleted
+	for _, i := range []int{0, 1, 4, 5, 6, 7} {
+		txID := txs[i].TxID
+		_, err := l2DB.GetTx(txID)
+		require.NoError(t, err)
+	}
+
+	// Query txs that have been deleted
+	for _, i := range []int{2, 3} {
+		txID := txs[i].TxID
+		_, err := l2DB.GetTx(txID)
+		require.Equal(t, sql.ErrNoRows, tracerr.Unwrap(err))
+	}
+}

@@ -34,6 +34,7 @@ RqAmount    *big.Int               `meddler:"rq_amount,bigintnull"`
 	RqFee       *common.FeeSelector    `meddler:"rq_fee"`
 	RqNonce     *common.Nonce          `meddler:"rq_nonce"`
 	Type        common.TxType          `meddler:"tx_type"`
+	ClientIP    string                 `meddler:"client_ip"`
 }
 
 // PoolTxAPI represents a L2 Tx pool with extra metadata used by the API

@@ -627,7 +627,9 @@ rq_token_id INT,
     rq_amount BYTEA,
     rq_fee SMALLINT,
     rq_nonce BIGINT,
-    tx_type VARCHAR(40) NOT NULL
+    tx_type VARCHAR(40) NOT NULL,
+    client_ip VARCHAR,
+    external_delete BOOLEAN NOT NULL DEFAULT false
 );
 
 -- +migrate StatementBegin

@@ -303,6 +303,7 @@ ForgeRetryInterval:     cfg.Coordinator.ForgeRetryInterval.Duration,
 				ForgeDelay:             cfg.Coordinator.ForgeDelay.Duration,
 				ForgeNoTxsDelay:        cfg.Coordinator.ForgeNoTxsDelay.Duration,
 				SyncRetryInterval:      cfg.Coordinator.SyncRetryInterval.Duration,
+				PurgeByExtDelInterval:  cfg.Coordinator.PurgeByExtDelInterval.Duration,
 				EthClientAttempts:      cfg.Coordinator.EthClient.Attempts,
 				EthClientAttemptsDelay: cfg.Coordinator.EthClient.AttemptsDelay.Duration,
 				EthNoReuseNonce:        cfg.Coordinator.EthClient.NoReuseNonce,