Package org.apache.hadoop.hbase.master.replication

Class SyncReplicationReplayWALManager

java.lang.Object

org.apache.hadoop.hbase.master.replication.SyncReplicationReplayWALManager

@Private
public class SyncReplicationReplayWALManager
extends Object

The manager for replaying remote wal.

First, it will be used to balance the replay work across all the region servers. We will record the region servers which have already been used for replaying wal, and prevent sending new replay work to it, until the previous replay work has been done, where we will remove the region server from the used worker set. See the comment for UsedReplayWorkersForPeer for more details.

Second, the logic for managing the remote wal directory is kept here. Before replaying the wals, we will rename the remote wal directory, the new name is called 'replay' directory, see renameToPeerReplayWALDir(String). This is used to prevent further writing of remote wals, which is very important for keeping consistency. And then we will start replaying all the wals, once a wal has been replayed, we will truncate the file, so that if there are crashes happen, we do not need to replay all the wals again, see finishReplayWAL(String) and isReplayWALFinished(String). After replaying all the wals, we will rename the 'replay' directory, the new name is called 'snapshot' directory. In the directory, we will keep all the names for the wals being replayed, since all the files should have been truncated. When we transitting original the ACTIVE cluster to STANDBY later, and there are region server crashes, we will see the wals in this directory to determine whether a wal should be split and replayed or not. You can see the code in SplitLogWorker for more details.

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
private static final class	SyncReplicationReplayWALManager.UsedReplayWorkersForPeer	This class is used to record the used workers(region servers) for a replication peer.

Field Summary

Fields

Modifier and Type	Field	Description
private final org.apache.hadoop.fs.FileSystem	fs
private static final org.slf4j.Logger	LOG
private final org.apache.hadoop.fs.Path	remoteWALDir
private final ServerManager	serverManager
private final ConcurrentMap<String,SyncReplicationReplayWALManager.UsedReplayWorkersForPeer>	usedWorkersByPeer
private final org.apache.hadoop.fs.Path	walRootDir

Constructor Summary

Constructors

Constructor	Description
SyncReplicationReplayWALManager(MasterServices services)

Method Summary

Modifier and Type	Method	Description
ServerName	acquirePeerWorker(String peerId, Procedure<?> proc)	Get a worker for replaying remote wal for a give peer.
void	addUsedPeerWorker(String peerId, ServerName worker)	Will only be called when loading procedures, where we need to construct the used worker set for each peer.
void	createPeerRemoteWALDir(String peerId)
private void	deleteDir(org.apache.hadoop.fs.Path dir, String peerId)
void	finishReplayWAL(String wal)
org.apache.hadoop.fs.Path	getRemoteWALDir()
List<org.apache.hadoop.fs.Path>	getReplayWALsAndCleanUpUnusedFiles(String peerId)
boolean	isReplayWALFinished(String wal)
void	registerPeer(String peerId)
void	releasePeerWorker(String peerId, ServerName worker, MasterProcedureScheduler scheduler)
void	removePeerRemoteWALs(String peerId)
String	removeWALRootPath(org.apache.hadoop.fs.Path path)
private void	rename(org.apache.hadoop.fs.Path src, org.apache.hadoop.fs.Path dst, String peerId)
void	renameToPeerReplayWALDir(String peerId)
void	renameToPeerSnapshotWALDir(String peerId)
void	unregisterPeer(String peerId)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Details

LOG

private static final org.slf4j.Logger LOG

serverManager

private final ServerManager serverManager

fs

private final org.apache.hadoop.fs.FileSystem fs

walRootDir

private final org.apache.hadoop.fs.Path walRootDir

remoteWALDir

private final org.apache.hadoop.fs.Path remoteWALDir

usedWorkersByPeer

private final ConcurrentMap<String,SyncReplicationReplayWALManager.UsedReplayWorkersForPeer> usedWorkersByPeer

Constructor Details

SyncReplicationReplayWALManager

public SyncReplicationReplayWALManager(MasterServices services)
                                throws IOException,
ReplicationException

Throws:

IOException

ReplicationException

Method Details

registerPeer

public void registerPeer(String peerId)

unregisterPeer

public void unregisterPeer(String peerId)

acquirePeerWorker

public ServerName acquirePeerWorker(String peerId,
 Procedure<?> proc)
                             throws ProcedureSuspendedException

Get a worker for replaying remote wal for a give peer. If no worker available, i.e, all the region servers have been used by others, a ProcedureSuspendedException will be thrown to suspend the procedure. And it will be woken up later when there are available workers, either by others release a worker, or there is a new region server joins the cluster.

Throws:

ProcedureSuspendedException

releasePeerWorker

public void releasePeerWorker(String peerId,
 ServerName worker,
 MasterProcedureScheduler scheduler)

addUsedPeerWorker

public void addUsedPeerWorker(String peerId,
 ServerName worker)

Will only be called when loading procedures, where we need to construct the used worker set for each peer.

createPeerRemoteWALDir

public void createPeerRemoteWALDir(String peerId)
                            throws IOException

Throws:

IOException

rename

private void rename(org.apache.hadoop.fs.Path src,
 org.apache.hadoop.fs.Path dst,
 String peerId)
             throws IOException

Throws:

IOException

renameToPeerReplayWALDir

public void renameToPeerReplayWALDir(String peerId)
                              throws IOException

Throws:

IOException

renameToPeerSnapshotWALDir

public void renameToPeerSnapshotWALDir(String peerId)
                                throws IOException

Throws:

IOException

getReplayWALsAndCleanUpUnusedFiles

public List<org.apache.hadoop.fs.Path> getReplayWALsAndCleanUpUnusedFiles(String peerId)
                                                                   throws IOException

Throws:

IOException

deleteDir

private void deleteDir(org.apache.hadoop.fs.Path dir,
 String peerId)
                throws IOException

Throws:

IOException

removePeerRemoteWALs

public void removePeerRemoteWALs(String peerId)
                          throws IOException

Throws:

IOException

removeWALRootPath

public String removeWALRootPath(org.apache.hadoop.fs.Path path)

finishReplayWAL

public void finishReplayWAL(String wal)
                     throws IOException

Throws:

IOException

isReplayWALFinished

public boolean isReplayWALFinished(String wal)
                            throws IOException

Throws:

IOException

getRemoteWALDir

public org.apache.hadoop.fs.Path getRemoteWALDir()