feat: pipelined dml has its own max-txn-ttl of 24 hours (#1224)

* feat: pipelined dml has its own max-txn-ttl of 24 hours

Signed-off-by: ekexium <eke@fastmail.com>

* fix gh actions

Signed-off-by: ekexium <eke@fastmail.com>

* style: fix lint

Signed-off-by: ekexium <eke@fastmail.com>

---------

Signed-off-by: ekexium <eke@fastmail.com>

.github/workflows/test.yml

@@ -48,3 +48,4 @@ jobs:
         uses: golangci/golangci-lint-action@v3
         with:
           version: v1.55.2
+          skip-pkg-cache: true

txnkv/transaction/2pc.go

@@ -81,6 +81,7 @@ type twoPhaseCommitAction interface {
 // Global variable set by config file.
 var (
 	ManagedLockTTL     uint64 = 20000               // 20s
+	MaxPipelinedTxnTTL uint64 = 24 * 60 * 60 * 1000 // 24h
 )
 
 var (
@@ -1145,7 +1146,7 @@ type ttlManager struct {
 	lockCtx *kv.LockCtx
 }
 
-func (tm *ttlManager) run(c *twoPhaseCommitter, lockCtx *kv.LockCtx) {
+func (tm *ttlManager) run(c *twoPhaseCommitter, lockCtx *kv.LockCtx, isPipelinedTxn bool) {
 	if _, err := util.EvalFailpoint("doNotKeepAlive"); err == nil {
 		return
 	}
@@ -1157,7 +1158,7 @@ func (tm *ttlManager) run(c *twoPhaseCommitter, lockCtx *kv.LockCtx) {
 	tm.ch = make(chan struct{})
 	tm.lockCtx = lockCtx
 
-	go keepAlive(c, tm.ch, c.primary(), lockCtx)
+	go keepAlive(c, tm.ch, c.primary(), lockCtx, isPipelinedTxn)
 }
 
 func (tm *ttlManager) close() {
@@ -1178,7 +1179,10 @@ const keepAliveMaxBackoff = 20000
 const pessimisticLockMaxBackoff = 20000
 const maxConsecutiveFailure = 10
 
-func keepAlive(c *twoPhaseCommitter, closeCh chan struct{}, primaryKey []byte, lockCtx *kv.LockCtx) {
+func keepAlive(
+	c *twoPhaseCommitter, closeCh chan struct{}, primaryKey []byte,
+	lockCtx *kv.LockCtx, isPipelinedTxn bool,
+) {
 	// Ticker is set to 1/2 of the ManagedLockTTL.
 	ticker := time.NewTicker(time.Duration(atomic.LoadUint64(&ManagedLockTTL)) * time.Millisecond / 2)
 	defer ticker.Stop()
@@ -1205,7 +1209,11 @@ func keepAlive(c *twoPhaseCommitter, closeCh chan struct{}, primaryKey []byte, l
 			}
 
 			uptime := uint64(oracle.ExtractPhysical(now) - oracle.ExtractPhysical(c.startTS))
-			if uptime > config.GetGlobalConfig().MaxTxnTTL {
+			maxTtl := config.GetGlobalConfig().MaxTxnTTL
+			if isPipelinedTxn {
+				maxTtl = max(maxTtl, MaxPipelinedTxnTTL)
+			}
+			if uptime > maxTtl {
 				// Checks maximum lifetime for the ttlManager, so when something goes wrong
 				// the key will not be locked forever.
 				logutil.Logger(bo.GetCtx()).Info("ttlManager live up to its lifetime",

txnkv/transaction/pessimistic.go

@@ -313,7 +313,7 @@ func (action actionPessimisticLock) handlePessimisticLockResponseNormalMode(
 		if batch.isPrimary {
 			// After locking the primary key, we should protect the primary lock from expiring
 			// now in case locking the remaining keys take a long time.
-			c.run(c, action.LockCtx)
+			c.run(c, action.LockCtx, false)
 		}
 
 		// Handle the case that the TiKV's version is too old and doesn't support `CheckExistence`.
@@ -412,7 +412,7 @@ func (action actionPessimisticLock) handlePessimisticLockResponseForceLockMode(
 		len(lockResp.Results) > 0 &&
 		lockResp.Results[0].Type != kvrpcpb.PessimisticLockKeyResultType_LockResultFailed {
 		// After locking the primary key, we should protect the primary lock from expiring.
-		c.run(c, action.LockCtx)
+		c.run(c, action.LockCtx, false)
 	}
 
 	if len(lockResp.Results) > 0 {

txnkv/transaction/pipelined_flush.go

@@ -198,7 +198,7 @@ func (action actionPipelinedFlush) handleSingleBatch(
 
 			if batch.isPrimary {
 				// start keepalive after primary key is written.
-				c.run(c, nil)
+				c.run(c, nil, true)
 			}
 			return nil
 		}

txnkv/transaction/prewrite.go

@@ -370,7 +370,7 @@ func (action actionPrewrite) handleSingleBatch(
 				// In this case 1PC is not expected to be used, but still check it for safety.
 				if int64(c.txnSize) > config.GetGlobalConfig().TiKVClient.TTLRefreshedTxnSize &&
 					prewriteResp.OnePcCommitTs == 0 {
-					c.run(c, nil)
+					c.run(c, nil, false)
 				}
 			}