ndb - bug#24664

  1) run lcp snapshot for both MM and DD tables (so I dont have to change restore to use WRITE)
  2) fix >= and > bug in lcp skip/keep handling
  3) very cool test prog for this :-)

storage/ndb/include/kernel/signaldata/BackupContinueB.hpp

@@ -33,7 +33,8 @@ private:
     BUFFER_FULL_FRAG_COMPLETE = 3,
     BUFFER_FULL_META  = 4,
     BACKUP_FRAGMENT_INFO = 5,
-    RESET_DISK_SPEED_COUNTER = 6
+    RESET_DISK_SPEED_COUNTER = 6,
+    ZDELAY_SCAN_NEXT = 7
   };
 };
 

storage/ndb/src/kernel/blocks/ERROR_codes.txt

@@ -8,7 +8,7 @@ Next DBDICT 6007
 Next DBDIH 7178
 Next DBTC 8039
 Next CMVMI 9000
-Next BACKUP 10036
+Next BACKUP 10038
 Next DBUTIL 11002
 Next DBTUX 12008
 Next SUMA 13001
@@ -425,6 +425,9 @@ Backup Stuff:
 10034: define backup reply error
 10035: Fail to allocate buffers
 
+10036: Halt backup for table >= 2
+10037: Resume backup (from 10036)
+
 11001: Send UTIL_SEQUENCE_REF (in master)
 
 5028:  Crash when receiving LQHKEYREQ (in non-master)

storage/ndb/src/kernel/blocks/backup/Backup.cpp

@@ -356,6 +356,25 @@ Backup::execCONTINUEB(Signal* signal)
 	       GetTabInfoReq::SignalLength, JBB);
     return;
   }
+  case BackupContinueB::ZDELAY_SCAN_NEXT:
+    if (ERROR_INSERTED(10036))
+    {
+      jam();
+      sendSignalWithDelay(BACKUP_REF, GSN_CONTINUEB, signal, 300, 
+			  signal->getLength());
+      return;
+    }
+    else
+    {
+      jam();
+      CLEAR_ERROR_INSERT_VALUE;
+      ndbout_c("Resuming backup");
+      memmove(signal->theData, signal->theData + 1, 
+	      4*ScanFragNextReq::SignalLength);
+      sendSignal(DBLQH_REF, GSN_SCAN_NEXTREQ, signal, 
+		 ScanFragNextReq::SignalLength, JBB);
+      return ;
+    }
   default:
     ndbrequire(0);
   }//switch
@@ -3920,6 +3939,22 @@ Backup::checkScan(Signal* signal, BackupFilePtr filePtr)
     req->transId2 = (BACKUP << 20) + (getOwnNodeId() << 8);
     req->batch_size_rows= 16;
     req->batch_size_bytes= 0;
+
+    if (ERROR_INSERTED(10036) && 
+	filePtr.p->tableId >= 2 &&
+	filePtr.p->operation.noOfRecords > 0)
+    {
+      ndbout_c("halting backup for table %d fragment: %d after %d records",
+	       filePtr.p->tableId,
+	       filePtr.p->fragmentNo,
+	       filePtr.p->operation.noOfRecords);
+      memmove(signal->theData+1, signal->theData, 
+	      4*ScanFragNextReq::SignalLength);
+      signal->theData[0] = BackupContinueB::ZDELAY_SCAN_NEXT;
+      sendSignalWithDelay(BACKUP_REF, GSN_CONTINUEB, signal, 
+			  300, 1+ScanFragNextReq::SignalLength);
+      return;
+    }
     if(ERROR_INSERTED(10032))
       sendSignalWithDelay(DBLQH_REF, GSN_SCAN_NEXTREQ, signal, 
 			  100, ScanFragNextReq::SignalLength);

storage/ndb/src/kernel/blocks/dbtup/DbtupCommit.cpp

@@ -152,10 +152,10 @@ void Dbtup::initOpConnection(Operationrec* regOperPtr)
 static
 inline
 bool
-operator>=(const Local_key& key1, const Local_key& key2)
+operator>(const Local_key& key1, const Local_key& key2)
 {
   return key1.m_page_no > key2.m_page_no ||
-    (key1.m_page_no == key2.m_page_no && key1.m_page_idx >= key2.m_page_idx);
+    (key1.m_page_no == key2.m_page_no && key1.m_page_idx > key2.m_page_idx);
 }
 
 void
@@ -187,7 +187,7 @@ Dbtup::dealloc_tuple(Signal* signal,
     Local_key rowid = regOperPtr->m_tuple_location;
     Local_key scanpos = scanOp.p->m_scanPos.m_key;
     rowid.m_page_no = page->frag_page_id;
-    if (rowid >= scanpos)
+    if (rowid > scanpos)
     {
       extra_bits = Tuple_header::LCP_KEEP; // Note REMOVE FREE
       ptr->m_operation_ptr_i = lcp_keep_list;
@@ -215,6 +215,7 @@ Dbtup::commit_operation(Signal* signal,
 {
   ndbassert(regOperPtr->op_struct.op_type != ZDELETE);
   
+  Uint32 lcpScan_ptr_i= regFragPtr->m_lcp_scan_op;
   Uint32 save= tuple_ptr->m_operation_ptr_i;
   Uint32 bits= tuple_ptr->m_header_bits;
 
@@ -264,7 +265,6 @@ Dbtup::commit_operation(Signal* signal,
     Local_key key;
     memcpy(&key, copy->get_disk_ref_ptr(regTabPtr), sizeof(Local_key));
     Uint32 logfile_group_id= regFragPtr->m_logfile_group_id;
-    Uint32 lcpScan_ptr_i= regFragPtr->m_lcp_scan_op;
 
     PagePtr diskPagePtr = *(PagePtr*)&m_pgman.m_ptr;
     ndbassert(diskPagePtr.p->m_page_no == key.m_page_no);
@@ -273,19 +273,6 @@ Dbtup::commit_operation(Signal* signal,
     if(copy_bits & Tuple_header::DISK_ALLOC)
     {
       disk_page_alloc(signal, regTabPtr, regFragPtr, &key, diskPagePtr, gci);
-
-      if(lcpScan_ptr_i != RNIL)
-      {
-	ScanOpPtr scanOp;
-	c_scanOpPool.getPtr(scanOp, lcpScan_ptr_i);
-	Local_key rowid = regOperPtr->m_tuple_location;
-	Local_key scanpos = scanOp.p->m_scanPos.m_key;
-	rowid.m_page_no = pagePtr.p->frag_page_id;
-	if(rowid >= scanpos)
-	{
-	  copy_bits |= Tuple_header::LCP_SKIP;
-	}
-      }
     }
     
     if(regTabPtr->m_attributes[DD].m_no_of_varsize == 0)
@@ -312,6 +299,18 @@ Dbtup::commit_operation(Signal* signal,
     copy_bits |= Tuple_header::DISK_PART;
   }
   
+  if(lcpScan_ptr_i != RNIL)
+  {
+    ScanOpPtr scanOp;
+    c_scanOpPool.getPtr(scanOp, lcpScan_ptr_i);
+    Local_key rowid = regOperPtr->m_tuple_location;
+    Local_key scanpos = scanOp.p->m_scanPos.m_key;
+    rowid.m_page_no = pagePtr.p->frag_page_id;
+    if(rowid > scanpos)
+    {
+       copy_bits |= Tuple_header::LCP_SKIP;
+    }
+  }
   
   Uint32 clear= 
     Tuple_header::ALLOC | Tuple_header::FREE |

storage/ndb/src/kernel/blocks/dbtup/DbtupScan.cpp

@@ -54,8 +54,7 @@ Dbtup::execACC_SCANREQ(Signal* signal)
     // flags
     Uint32 bits = 0;
     
-    if (!AccScanReq::getLcpScanFlag(req->requestInfo) ||
-	tablePtr.p->m_no_of_disk_attributes == 0)
+    if (!AccScanReq::getLcpScanFlag(req->requestInfo))
     {
       // seize from pool and link to per-fragment list
       LocalDLList<ScanOp> list(c_scanOpPool, frag.m_scanList);
@@ -1052,24 +1051,21 @@ Dbtup::execLCP_FRAG_ORD(Signal* signal)
   tablePtr.i = req->tableId;
   ptrCheckGuard(tablePtr, cnoOfTablerec, tablerec);
 
-  if(tablePtr.p->m_no_of_disk_attributes)
-  {
-    jam();
-    FragrecordPtr fragPtr;
-    Uint32 fragId = req->fragmentId;
-    fragPtr.i = RNIL;
-    getFragmentrec(fragPtr, fragId, tablePtr.p);
-    ndbrequire(fragPtr.i != RNIL);
-    Fragrecord& frag = *fragPtr.p;
-    
-    ndbrequire(frag.m_lcp_scan_op == RNIL && c_lcp_scan_op != RNIL);
-    frag.m_lcp_scan_op = c_lcp_scan_op;
-    ScanOpPtr scanPtr;
-    c_scanOpPool.getPtr(scanPtr, frag.m_lcp_scan_op);
-    ndbrequire(scanPtr.p->m_fragPtrI == RNIL);
-    scanPtr.p->m_fragPtrI = fragPtr.i;
-    
-    scanFirst(signal, scanPtr);
-    scanPtr.p->m_state = ScanOp::First;
-  }
+  jam();
+  FragrecordPtr fragPtr;
+  Uint32 fragId = req->fragmentId;
+  fragPtr.i = RNIL;
+  getFragmentrec(fragPtr, fragId, tablePtr.p);
+  ndbrequire(fragPtr.i != RNIL);
+  Fragrecord& frag = *fragPtr.p;
+  
+  ndbrequire(frag.m_lcp_scan_op == RNIL && c_lcp_scan_op != RNIL);
+  frag.m_lcp_scan_op = c_lcp_scan_op;
+  ScanOpPtr scanPtr;
+  c_scanOpPool.getPtr(scanPtr, frag.m_lcp_scan_op);
+  ndbrequire(scanPtr.p->m_fragPtrI == RNIL);
+  scanPtr.p->m_fragPtrI = fragPtr.i;
+  
+  scanFirst(signal, scanPtr);
+  scanPtr.p->m_state = ScanOp::First;
 }

storage/ndb/test/ndbapi/testSystemRestart.cpp

@@ -1162,6 +1162,64 @@ runBug21536(NDBT_Context* ctx, NDBT_Step* step)
   return result;
 }
 
+int 
+runBug24664(NDBT_Context* ctx, NDBT_Step* step)
+{
+  int result = NDBT_OK;
+  NdbRestarter restarter;
+  Ndb* pNdb = GETNDB(step);
+  const Uint32 nodeCount = restarter.getNumDbNodes();
+
+  int records = ctx->getNumRecords();
+  UtilTransactions utilTrans(*ctx->getTab());
+  HugoTransactions hugoTrans(*ctx->getTab());
+
+  int args[] = { DumpStateOrd::DihMaxTimeBetweenLCP };
+  int dump[] = { DumpStateOrd::DihStartLcpImmediately };
+  
+  int filter[] = { 15, NDB_MGM_EVENT_CATEGORY_CHECKPOINT, 0 };
+  NdbLogEventHandle handle = 
+    ndb_mgm_create_logevent_handle(restarter.handle, filter);
+
+  struct ndb_logevent event;
+
+  do {
+    CHECK(restarter.dumpStateAllNodes(args, 1) == 0);
+    CHECK(restarter.dumpStateAllNodes(dump, 1) == 0);
+    while(ndb_logevent_get_next(handle, &event, 0) >= 0 &&
+	  event.type != NDB_LE_LocalCheckpointStarted);
+    while(ndb_logevent_get_next(handle, &event, 0) >= 0 &&
+	  event.type != NDB_LE_LocalCheckpointCompleted);
+    
+    if (hugoTrans.loadTable(GETNDB(step), records) != 0){
+      return NDBT_FAILED;
+    }
+  
+    restarter.insertErrorInAllNodes(10036); // Hang LCP
+    CHECK(restarter.dumpStateAllNodes(dump, 1) == 0);
+    while(ndb_logevent_get_next(handle, &event, 0) >= 0 &&
+	  event.type != NDB_LE_LocalCheckpointStarted);
+    NdbSleep_SecSleep(3);
+    CHECK(utilTrans.clearTable(pNdb,  records) == 0);
+    if (hugoTrans.loadTable(GETNDB(step), records) != 0){
+      return NDBT_FAILED;
+    }
+
+    restarter.insertErrorInAllNodes(10037); // Resume LCP
+    while(ndb_logevent_get_next(handle, &event, 0) >= 0 &&
+	  event.type != NDB_LE_LocalCheckpointCompleted);
+
+    while(ndb_logevent_get_next(handle, &event, 0) >= 0 &&
+	  event.type != NDB_LE_GlobalCheckpointCompleted);
+    while(ndb_logevent_get_next(handle, &event, 0) >= 0 &&
+	  event.type != NDB_LE_GlobalCheckpointCompleted);
+    restarter.restartAll(false, false, true);
+    CHECK(restarter.waitClusterStarted() == 0);
+  } while(false);
+  
+  return result;
+}
+
 NDBT_TESTSUITE(testSystemRestart);
 TESTCASE("SR1", 
 	 "Basic system restart test. Focus on testing restart from REDO log.\n"
@@ -1334,6 +1392,14 @@ TESTCASE("Bug21536",
   STEP(runBug21536);
   FINALIZER(runClearTable);
 }
+TESTCASE("Bug24664",
+	 "Check handling of LCP skip/keep")
+{
+  INITIALIZER(runWaitStarted);
+  INITIALIZER(runClearTable);
+  STEP(runBug24664);
+  FINALIZER(runClearTable);
+}
 NDBT_TESTSUITE_END(testSystemRestart);
 
 int main(int argc, const char** argv){

storage/ndb/test/run-test/daily-basic-tests.txt

@@ -752,6 +752,10 @@ max-time: 300
 cmd: testNodeRestart
 args: -n Bug24543 T1
 
+max-time: 1500
+cmd: testSystemRestart
+args: -n Bug24664
+
 # OLD FLEX
 max-time: 500
 cmd: flexBench