/// <exception cref="System.IO.IOException"/>
internal static void DoMerge(CheckpointSignature sig, RemoteEditLogManifest manifest
, bool loadImage, FSImage dstImage, FSNamesystem dstNamesystem)
{
NNStorage dstStorage = dstImage.GetStorage();
dstStorage.SetStorageInfo(sig);
if (loadImage)
{
FilePath file = dstStorage.FindImageFile(NNStorage.NameNodeFile.Image, sig.mostRecentCheckpointTxId
);
if (file == null)
{
throw new IOException("Couldn't find image file at txid " + sig.mostRecentCheckpointTxId
+ " even though it should have " + "just been downloaded");
}
dstNamesystem.WriteLock();
try
{
dstImage.ReloadFromImageFile(file, dstNamesystem);
}
finally
{
dstNamesystem.WriteUnlock();
}
dstNamesystem.ImageLoadComplete();
}
// error simulation code for junit test
CheckpointFaultInjector.GetInstance().DuringMerge();
Checkpointer.RollForwardByApplyingLogs(manifest, dstImage, dstNamesystem);
// The following has the side effect of purging old fsimages/edit logs.
dstImage.SaveFSImageInAllDirs(dstNamesystem, dstImage.GetLastAppliedTxId());
dstStorage.WriteAll();
}
public _PrivilegedExceptionAction_444(FSImage dstImage, CheckpointSignature sig,
Uri nnHostPort, RemoteEditLogManifest manifest)
{
this.dstImage = dstImage;
this.sig = sig;
this.nnHostPort = nnHostPort;
this.manifest = manifest;
}
public virtual void TestEditLogRolling()
{
// start a cluster
Configuration conf = new HdfsConfiguration();
MiniDFSCluster cluster = null;
FileSystem fileSys = null;
AtomicReference <Exception> caughtErr = new AtomicReference <Exception>();
try
{
cluster = new MiniDFSCluster.Builder(conf).NumDataNodes(NumDataNodes).Build();
cluster.WaitActive();
fileSys = cluster.GetFileSystem();
NamenodeProtocols nn = cluster.GetNameNode().GetRpcServer();
FSImage fsimage = cluster.GetNamesystem().GetFSImage();
Storage.StorageDirectory sd = fsimage.GetStorage().GetStorageDir(0);
StartTransactionWorkers(nn, caughtErr);
long previousLogTxId = 1;
for (int i = 0; i < NumRolls && caughtErr.Get() == null; i++)
{
try
{
Sharpen.Thread.Sleep(20);
}
catch (Exception)
{
}
Log.Info("Starting roll " + i + ".");
CheckpointSignature sig = nn.RollEditLog();
long nextLog = sig.curSegmentTxId;
string logFileName = NNStorage.GetFinalizedEditsFileName(previousLogTxId, nextLog
- 1);
previousLogTxId += VerifyEditLogs(cluster.GetNamesystem(), fsimage, logFileName,
previousLogTxId);
NUnit.Framework.Assert.AreEqual(previousLogTxId, nextLog);
FilePath expectedLog = NNStorage.GetInProgressEditsFile(sd, previousLogTxId);
NUnit.Framework.Assert.IsTrue("Expect " + expectedLog + " to exist", expectedLog.
Exists());
}
}
finally
{
StopTransactionWorkers();
if (caughtErr.Get() != null)
{
throw new RuntimeException(caughtErr.Get());
}
if (fileSys != null)
{
fileSys.Close();
}
if (cluster != null)
{
cluster.Shutdown();
}
}
}
/// <summary>Get edits filename</summary>
/// <returns>edits file name for cluster</returns>
/// <exception cref="System.IO.IOException"/>
private string GetEditsFilename(CheckpointSignature sig)
{
FSImage image = cluster.GetNameNode().GetFSImage();
// it was set up to only have ONE StorageDirectory
IEnumerator <Storage.StorageDirectory> it = image.GetStorage().DirIterator(NNStorage.NameNodeDirType
.Edits);
Storage.StorageDirectory sd = it.Next();
FilePath ret = NNStorage.GetFinalizedEditsFile(sd, 1, sig.curSegmentTxId - 1);
System.Diagnostics.Debug.Assert(ret.Exists(), "expected " + ret + " exists");
return(ret.GetAbsolutePath());
}
/// <exception cref="System.Exception"/>
public virtual void TestSaveWhileEditsRolled()
{
Configuration conf = GetConf();
NameNode.InitMetrics(conf, HdfsServerConstants.NamenodeRole.Namenode);
DFSTestUtil.FormatNameNode(conf);
FSNamesystem fsn = FSNamesystem.LoadFromDisk(conf);
try
{
DoAnEdit(fsn, 1);
CheckpointSignature sig = fsn.RollEditLog();
Log.Warn("Checkpoint signature: " + sig);
// Do another edit
DoAnEdit(fsn, 2);
// Save namespace
fsn.SetSafeMode(HdfsConstants.SafeModeAction.SafemodeEnter);
fsn.SaveNamespace();
// Now shut down and restart the NN
fsn.Close();
fsn = null;
// Start a new namesystem, which should be able to recover
// the namespace from the previous incarnation.
fsn = FSNamesystem.LoadFromDisk(conf);
// Make sure the image loaded including our edits.
CheckEditExists(fsn, 1);
CheckEditExists(fsn, 2);
}
finally
{
if (fsn != null)
{
fsn.Close();
}
}
}
/// <summary>Generates edits with all op codes and returns the edits filename</summary>
/// <exception cref="System.IO.IOException"/>
public virtual string GenerateEdits()
{
CheckpointSignature signature = RunOperations();
return(GetEditsFilename(signature));
}
/// <summary>Create a new checkpoint</summary>
/// <exception cref="System.IO.IOException"/>
internal virtual void DoCheckpoint()
{
BackupImage bnImage = GetFSImage();
NNStorage bnStorage = bnImage.GetStorage();
long startTime = Time.MonotonicNow();
bnImage.FreezeNamespaceAtNextRoll();
NamenodeCommand cmd = GetRemoteNamenodeProxy().StartCheckpoint(backupNode.GetRegistration
());
CheckpointCommand cpCmd = null;
switch (cmd.GetAction())
{
case NamenodeProtocol.ActShutdown:
{
Shutdown();
throw new IOException("Name-node " + backupNode.nnRpcAddress + " requested shutdown."
);
}
case NamenodeProtocol.ActCheckpoint:
{
cpCmd = (CheckpointCommand)cmd;
break;
}
default:
{
throw new IOException("Unsupported NamenodeCommand: " + cmd.GetAction());
}
}
bnImage.WaitUntilNamespaceFrozen();
CheckpointSignature sig = cpCmd.GetSignature();
// Make sure we're talking to the same NN!
sig.ValidateStorageInfo(bnImage);
long lastApplied = bnImage.GetLastAppliedTxId();
Log.Debug("Doing checkpoint. Last applied: " + lastApplied);
RemoteEditLogManifest manifest = GetRemoteNamenodeProxy().GetEditLogManifest(bnImage
.GetLastAppliedTxId() + 1);
bool needReloadImage = false;
if (!manifest.GetLogs().IsEmpty())
{
RemoteEditLog firstRemoteLog = manifest.GetLogs()[0];
// we don't have enough logs to roll forward using only logs. Need
// to download and load the image.
if (firstRemoteLog.GetStartTxId() > lastApplied + 1)
{
Log.Info("Unable to roll forward using only logs. Downloading " + "image with txid "
+ sig.mostRecentCheckpointTxId);
MD5Hash downloadedHash = TransferFsImage.DownloadImageToStorage(backupNode.nnHttpAddress
, sig.mostRecentCheckpointTxId, bnStorage, true);
bnImage.SaveDigestAndRenameCheckpointImage(NNStorage.NameNodeFile.Image, sig.mostRecentCheckpointTxId
, downloadedHash);
lastApplied = sig.mostRecentCheckpointTxId;
needReloadImage = true;
}
if (firstRemoteLog.GetStartTxId() > lastApplied + 1)
{
throw new IOException("No logs to roll forward from " + lastApplied);
}
// get edits files
foreach (RemoteEditLog log in manifest.GetLogs())
{
TransferFsImage.DownloadEditsToStorage(backupNode.nnHttpAddress, log, bnStorage);
}
if (needReloadImage)
{
Log.Info("Loading image with txid " + sig.mostRecentCheckpointTxId);
FilePath file = bnStorage.FindImageFile(NNStorage.NameNodeFile.Image, sig.mostRecentCheckpointTxId
);
bnImage.ReloadFromImageFile(file, backupNode.GetNamesystem());
}
RollForwardByApplyingLogs(manifest, bnImage, backupNode.GetNamesystem());
}
long txid = bnImage.GetLastAppliedTxId();
backupNode.namesystem.WriteLock();
try
{
backupNode.namesystem.SetImageLoaded();
if (backupNode.namesystem.GetBlocksTotal() > 0)
{
backupNode.namesystem.SetBlockTotal();
}
bnImage.SaveFSImageInAllDirs(backupNode.GetNamesystem(), txid);
bnStorage.WriteAll();
}
finally
{
backupNode.namesystem.WriteUnlock();
}
if (cpCmd.NeedToReturnImage())
{
TransferFsImage.UploadImageFromStorage(backupNode.nnHttpAddress, conf, bnStorage,
NNStorage.NameNodeFile.Image, txid);
}
GetRemoteNamenodeProxy().EndCheckpoint(backupNode.GetRegistration(), sig);
if (backupNode.GetRole() == HdfsServerConstants.NamenodeRole.Backup)
{
bnImage.ConvergeJournalSpool();
}
backupNode.SetRegistration();
// keep registration up to date
long imageSize = bnImage.GetStorage().GetFsImageName(txid).Length();
Log.Info("Checkpoint completed in " + (Time.MonotonicNow() - startTime) / 1000 +
" seconds." + " New Image Size: " + imageSize);
}
public virtual bool DoCheckpoint()
{
checkpointImage.EnsureCurrentDirExists();
NNStorage dstStorage = checkpointImage.GetStorage();
// Tell the namenode to start logging transactions in a new edit file
// Returns a token that would be used to upload the merged image.
CheckpointSignature sig = namenode.RollEditLog();
bool loadImage = false;
bool isFreshCheckpointer = (checkpointImage.GetNamespaceID() == 0);
bool isSameCluster = (dstStorage.VersionSupportsFederation(NameNodeLayoutVersion.
Features) && sig.IsSameCluster(checkpointImage)) || (!dstStorage.VersionSupportsFederation
(NameNodeLayoutVersion.Features) && sig.NamespaceIdMatches(checkpointImage));
if (isFreshCheckpointer || (isSameCluster && !sig.StorageVersionMatches(checkpointImage
.GetStorage())))
{
// if we're a fresh 2NN, or if we're on the same cluster and our storage
// needs an upgrade, just take the storage info from the server.
dstStorage.SetStorageInfo(sig);
dstStorage.SetClusterID(sig.GetClusterID());
dstStorage.SetBlockPoolID(sig.GetBlockpoolID());
loadImage = true;
}
sig.ValidateStorageInfo(checkpointImage);
// error simulation code for junit test
CheckpointFaultInjector.GetInstance().AfterSecondaryCallsRollEditLog();
RemoteEditLogManifest manifest = namenode.GetEditLogManifest(sig.mostRecentCheckpointTxId
+ 1);
// Fetch fsimage and edits. Reload the image if previous merge failed.
loadImage |= DownloadCheckpointFiles(fsName, checkpointImage, sig, manifest) | checkpointImage
.HasMergeError();
try
{
DoMerge(sig, manifest, loadImage, checkpointImage, namesystem);
}
catch (IOException ioe)
{
// A merge error occurred. The in-memory file system state may be
// inconsistent, so the image and edits need to be reloaded.
checkpointImage.SetMergeError();
throw;
}
// Clear any error since merge was successful.
checkpointImage.ClearMergeError();
//
// Upload the new image into the NameNode. Then tell the Namenode
// to make this new uploaded image as the most current image.
//
long txid = checkpointImage.GetLastAppliedTxId();
TransferFsImage.UploadImageFromStorage(fsName, conf, dstStorage, NNStorage.NameNodeFile
.Image, txid);
// error simulation code for junit test
CheckpointFaultInjector.GetInstance().AfterSecondaryUploadsNewImage();
Log.Warn("Checkpoint done. New Image Size: " + dstStorage.GetFsImageName(txid).Length
());
if (legacyOivImageDir != null && !legacyOivImageDir.IsEmpty())
{
try
{
checkpointImage.SaveLegacyOIVImage(namesystem, legacyOivImageDir, new Canceler());
}
catch (IOException e)
{
Log.Warn("Failed to write legacy OIV image: ", e);
}
}
return(loadImage);
}
/// <summary>
/// Download <code>fsimage</code> and <code>edits</code>
/// files from the name-node.
/// </summary>
/// <returns>true if a new image has been downloaded and needs to be loaded</returns>
/// <exception cref="System.IO.IOException"/>
internal static bool DownloadCheckpointFiles(Uri nnHostPort, FSImage dstImage, CheckpointSignature
sig, RemoteEditLogManifest manifest)
{
// Sanity check manifest - these could happen if, eg, someone on the
// NN side accidentally rmed the storage directories
if (manifest.GetLogs().IsEmpty())
{
throw new IOException("Found no edit logs to download on NN since txid " + sig.mostRecentCheckpointTxId
);
}
long expectedTxId = sig.mostRecentCheckpointTxId + 1;
if (manifest.GetLogs()[0].GetStartTxId() != expectedTxId)
{
throw new IOException("Bad edit log manifest (expected txid = " + expectedTxId +
": " + manifest);
}
try
{
bool b = UserGroupInformation.GetCurrentUser().DoAs(new _PrivilegedExceptionAction_444
(dstImage, sig, nnHostPort, manifest));
// get fsimage
// get edits file
// true if we haven't loaded all the transactions represented by the
// downloaded fsimage.
return(b);
}
catch (Exception e)
{
throw new RuntimeException(e);
}
}
