bug#9924 - ndb backup abort handling

  Redo abort handling according to descr. in Backup.txt
bug#9960 - ndb backup
      increase wait completed timeout to 48 hours


ndb/include/kernel/signaldata/BackupImpl.hpp:
  Add nodeid to reply to be able to fake reply during NF
ndb/include/kernel/signaldata/BackupSignalData.hpp:
  new error codes
ndb/src/common/debugger/signaldata/BackupImpl.cpp:
  fix printout
ndb/src/kernel/blocks/backup/Backup.cpp:
  bug#9924 - ndb backup abort handling
    Redo abort handling according to descr. in Backup.txt
ndb/src/kernel/blocks/backup/Backup.hpp:
  bug#9924 - ndb backup abort handling
    Redo abort handling according to descr. in Backup.txt
ndb/src/kernel/blocks/backup/Backup.txt:
  bug#9924 - ndb backup abort handling
    Redo abort handling according to descr. in Backup.txt
ndb/src/kernel/blocks/backup/BackupInit.cpp:
  bug#9924 - ndb backup abort handling
    Redo abort handling according to descr. in Backup.txt
ndb/src/kernel/blocks/cmvmi/Cmvmi.cpp:
  Init own version
ndb/src/mgmapi/mgmapi.cpp:
  bug#9960 - ndb backup
    increase wait completed timeout to 48 hours
ndb/src/mgmsrv/MgmtSrvr.cpp:
  Handle node failures activly
    (mainly for backup...)
ndb/src/mgmsrv/MgmtSrvr.hpp:
  Handle node failures activly
    (mainly for backup...)
ndb/src/mgmsrv/MgmtSrvrGeneralSignalHandling.cpp:
  Handle node failures activly
    (mainly for backup...)
ndb/src/ndbapi/ndberror.c:
  new error codes
ndb/test/ndbapi/testBackup.cpp:
  fix return codes
ndb/test/run-test/daily-basic-tests.txt:
  Add failure test cases to autotest
ndb/test/src/NdbBackup.cpp:
  fix error codes
  introduce checking of backup resources after each test

3 files changed, 18 insertions, 21 deletions

diff --git a/ndb/src/mgmsrv/MgmtSrvr.cpp b/ndb/src/mgmsrv/MgmtSrvr.cpp
index fb05e57e138..ceaedc9955b 100644
--- a/ndb/src/mgmsrv/MgmtSrvr.cpp
+++ b/ndb/src/mgmsrv/MgmtSrvr.cpp
@@ -791,7 +791,7 @@ MgmtSrvr::restartNode(int processId, bool nostart,
     result = sendSignal(processId, NO_WAIT, signal, true);
   }
   
-  if (result == -1) {
+  if (result == -1 && theWaitState != WAIT_NODEFAILURE) {
     m_stopRec.inUse = false;
     return SEND_OR_RECEIVE_FAILED;
   }
@@ -1920,6 +1920,7 @@ MgmtSrvr::handleReceivedSignal(NdbApiSignal* signal)
 #ifdef VM_TRACE
       ndbout_c("I'm not master resending to %d", aNodeId);
 #endif
+      theWaitNode= aNodeId;
       NdbApiSignal aSignal(_ownReference);
       BackupReq* req = CAST_PTR(BackupReq, aSignal.getDataPtrSend());
       aSignal.set(TestOrd::TraceAPI, BACKUP, GSN_BACKUP_REQ, 
@@ -1947,6 +1948,7 @@ MgmtSrvr::handleReceivedSignal(NdbApiSignal* signal)
     event.Event = BackupEvent::BackupAborted;
     event.Aborted.Reason = rep->reason;
     event.Aborted.BackupId = rep->backupId;
+    event.Aborted.ErrorCode = rep->reason;
     backupCallback(event);
   }
   break;
@@ -2076,6 +2078,13 @@ MgmtSrvr::handleStatus(NodeId nodeId, bool alive, bool nfComplete)
       handleStopReply(nodeId, 0);
       DBUG_VOID_RETURN;
     }
+
+    if(theWaitNode == nodeId && 
+       theWaitState != NO_WAIT && theWaitState != WAIT_STOP)
+    {
+      theWaitState = WAIT_NODEFAILURE;
+      NdbCondition_Signal(theMgmtWaitForResponseCondPtr);
+    }
   }
   
   eventReport(_ownNodeId, theData);
@@ -2427,7 +2436,7 @@ MgmtSrvr::startBackup(Uint32& backupId, int waitCompleted)
   int result;
   if (waitCompleted == 2) {
     result = sendRecSignal(nodeId, WAIT_BACKUP_COMPLETED,
-			   signal, true, 30*60*1000 /*30 secs*/);
+			   signal, true, 48*60*60*1000 /* 48 hours */);
   }
   else if (waitCompleted == 1) {
     result = sendRecSignal(nodeId, WAIT_BACKUP_STARTED,
@@ -2456,22 +2465,6 @@ MgmtSrvr::startBackup(Uint32& backupId, int waitCompleted)
       return -1;
       break;
     }
-  } else {
-    switch(m_lastBackupEvent.Event){
-    case BackupEvent::BackupCompleted:
-      backupId = m_lastBackupEvent.Completed.BackupId;
-      break;
-    case BackupEvent::BackupStarted:
-      backupId = m_lastBackupEvent.Started.BackupId;
-      break;
-    case BackupEvent::BackupFailedToStart:
-      return m_lastBackupEvent.FailedToStart.ErrorCode;
-    case BackupEvent::BackupAborted:
-      return m_lastBackupEvent.Aborted.ErrorCode;
-    default:
-      return -1;
-      break;
-    }
   }
   
   return 0;
diff --git a/ndb/src/mgmsrv/MgmtSrvr.hpp b/ndb/src/mgmsrv/MgmtSrvr.hpp
index a05b29b7f31..ce78983b3c3 100644
--- a/ndb/src/mgmsrv/MgmtSrvr.hpp
+++ b/ndb/src/mgmsrv/MgmtSrvr.hpp
@@ -611,7 +611,8 @@ private:
     WAIT_STOP,
     WAIT_BACKUP_STARTED,
     WAIT_BACKUP_COMPLETED,
-    WAIT_VERSION
+    WAIT_VERSION,
+    WAIT_NODEFAILURE
   };
 
   /**
@@ -695,6 +696,7 @@ private:
   NdbApiSignal* theSignalIdleList;
   // List of unused signals
   
+  Uint32 theWaitNode;
   WaitSignalType theWaitState;
   // State denoting a set of signals we accept to recieve.
 
diff --git a/ndb/src/mgmsrv/MgmtSrvrGeneralSignalHandling.cpp b/ndb/src/mgmsrv/MgmtSrvrGeneralSignalHandling.cpp
index 2126c9d358d..f93948abc75 100644
--- a/ndb/src/mgmsrv/MgmtSrvrGeneralSignalHandling.cpp
+++ b/ndb/src/mgmsrv/MgmtSrvrGeneralSignalHandling.cpp
@@ -108,6 +108,7 @@ MgmtSrvr::sendRecSignal(Uint16 aNodeId,
     return -1;
   }
   theWaitState = aWaitState;
+  theWaitNode = aNodeId;
   return receiveOptimisedResponse(waitTime);
 }
 
@@ -119,11 +120,12 @@ MgmtSrvr::receiveOptimisedResponse(int waitTime)
   theFacade->checkForceSend(_blockNumber);
   NDB_TICKS maxTime = NdbTick_CurrentMillisecond() + waitTime;
   
-  while (theWaitState != NO_WAIT && waitTime > 0) {
+  while (theWaitState != NO_WAIT && theWaitState != WAIT_NODEFAILURE
+	 && waitTime > 0) {
     NdbCondition_WaitTimeout(theMgmtWaitForResponseCondPtr, 
 			     theFacade->theMutexPtr,
 			     waitTime);
-    if(theWaitState == NO_WAIT)
+    if(theWaitState == NO_WAIT || theWaitState == WAIT_NODEFAILURE)
       break;
     waitTime = (maxTime - NdbTick_CurrentMillisecond());
   }//while