org.apache.hadoop.hbase.master
Class SplitLogManager

java.lang.Object
  org.apache.hadoop.hbase.zookeeper.ZooKeeperListener
      org.apache.hadoop.hbase.master.SplitLogManager

@InterfaceAudience.Private
public class SplitLogManager
extends ZooKeeperListener

Distributes the task of log splitting to the available region servers. Coordination happens via zookeeper. For every log file that has to be split a znode is created under /hbase/splitlog. SplitLogWorkers race to grab a task.

SplitLogManager monitors the task znodes that it creates using the timeoutMonitor thread. If a task's progress is slow then resubmit(String, Task, ResubmitDirective) will take away the task from the owner SplitLogWorker and the task will be up for grabs again. When the task is done then the task's znode is deleted by SplitLogManager.

Clients call splitLogDistributed(Path) to split a region server's log files. The caller thread waits in this method until all the log files have been split.

All the zookeeper calls made by this class are asynchronous. This is mainly to help reduce response time seen by the callers.

There is race in this design between the SplitLogManager and the SplitLogWorker. SplitLogManager might re-queue a task that has in reality already been completed by a SplitLogWorker. We rely on the idempotency of the log splitting task for correctness.

It is also assumed that every log splitting task is unique and once completed (either with success or with error) it will be not be submitted again. If a task is resubmitted then there is a risk that old "delete task" can delete the re-submission.

Nested Class Summary

static interface

SplitLogManager.TaskFinisher SplitLogManager can use objects implementing this interface to finish off a partially done task by SplitLogWorker.

Field Summary

static int	DEFAULT_MAX_RESUBMIT
static int	DEFAULT_TIMEOUT
static int	DEFAULT_UNASSIGNED_TIMEOUT
static int	DEFAULT_ZK_RETRIES
boolean	ignoreZKDeleteForTesting
protected ReentrantLock	recoveringRegionLock In distributedLogReplay mode, we need touch both splitlog and recovering-regions znodes in one operation.

Fields inherited from class org.apache.hadoop.hbase.zookeeper.ZooKeeperListener

watcher

Constructor Summary

SplitLogManager(ZooKeeperWatcher zkw, org.apache.hadoop.conf.Configuration conf, Stoppable stopper, MasterServices master, ServerName serverName, boolean masterRecovery)
Wrapper around SplitLogManager(ZooKeeperWatcher zkw, Configuration conf, Stoppable stopper, MasterServices master, ServerName serverName, boolean masterRecovery, TaskFinisher tf) that provides a task finisher for copying recovered edits to their final destination.

SplitLogManager(ZooKeeperWatcher zkw, org.apache.hadoop.conf.Configuration conf, Stoppable stopper, MasterServices master, ServerName serverName, boolean masterRecovery, SplitLogManager.TaskFinisher tf)
Its OK to construct this object even when region-servers are not online.

Method Summary

static void	deleteRecoveringRegionZNodes(ZooKeeperWatcher watcher, List<String> regions)
org.apache.hadoop.hbase.protobuf.generated.ZooKeeperProtos.SplitLogTask.RecoveryMode	getRecoveryMode()
static org.apache.hadoop.hbase.protobuf.generated.ZooKeeperProtos.RegionStoreSequenceIds	getRegionFlushedSequenceId(ZooKeeperWatcher zkw, String serverName, String encodedRegionName) This function is used in distributedLogReplay to fetch last flushed sequence id from ZK
static boolean	isRegionMarkedRecoveringInZK(ZooKeeperWatcher zkw, String regionEncodedName) check if /hbase/recovering-regions/ exists.
void	nodeDataChanged(String path)
static long	parseLastFlushedSequenceIdFrom(byte[] bytes)
void	setRecoveryMode(boolean isForInitialization) This function is to set recovery mode from outstanding split log tasks from before or current configuration setting
long	splitLogDistributed(List<org.apache.hadoop.fs.Path> logDirs) The caller will block until all the log files of the given region server have been processed - successfully split or an error is encountered - by an available worker region server.
long	splitLogDistributed(org.apache.hadoop.fs.Path logDir)
long	splitLogDistributed(Set<ServerName> serverNames, List<org.apache.hadoop.fs.Path> logDirs, org.apache.hadoop.fs.PathFilter filter) The caller will block until all the hbase:meta log files of the given region server have been processed - successfully split or an error is encountered - by an available worker region server.
void	stop()

Methods inherited from class org.apache.hadoop.hbase.zookeeper.ZooKeeperListener

getWatcher, nodeChildrenChanged, nodeCreated, nodeDeleted

Methods inherited from class java.lang.Object

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

Field Detail

DEFAULT_TIMEOUT

public static final int DEFAULT_TIMEOUT

See Also:

Constant Field Values

DEFAULT_ZK_RETRIES

public static final int DEFAULT_ZK_RETRIES

See Also:

Constant Field Values

DEFAULT_MAX_RESUBMIT

public static final int DEFAULT_MAX_RESUBMIT

See Also:

Constant Field Values

DEFAULT_UNASSIGNED_TIMEOUT

public static final int DEFAULT_UNASSIGNED_TIMEOUT

See Also:

Constant Field Values

ignoreZKDeleteForTesting

public boolean ignoreZKDeleteForTesting

recoveringRegionLock

protected final ReentrantLock recoveringRegionLock

In distributedLogReplay mode, we need touch both splitlog and recovering-regions znodes in one operation. So the lock is used to guard such cases.

Constructor Detail

SplitLogManager

public SplitLogManager(ZooKeeperWatcher zkw,
                       org.apache.hadoop.conf.Configuration conf,
                       Stoppable stopper,
                       MasterServices master,
                       ServerName serverName,
                       boolean masterRecovery)
                throws InterruptedIOException,
                       org.apache.zookeeper.KeeperException

Wrapper around SplitLogManager(ZooKeeperWatcher zkw, Configuration conf, Stoppable stopper, MasterServices master, ServerName serverName, boolean masterRecovery, TaskFinisher tf) that provides a task finisher for copying recovered edits to their final destination. The task finisher has to be robust because it can be arbitrarily restarted or called multiple times.

Parameters:

zkw - the ZK watcher

conf - the HBase configuration

stopper - the stoppable in case anything is wrong

master - the master services

serverName - the master server name

masterRecovery - an indication if the master is in recovery

Throws:

org.apache.zookeeper.KeeperException

InterruptedIOException

SplitLogManager

public SplitLogManager(ZooKeeperWatcher zkw,
                       org.apache.hadoop.conf.Configuration conf,
                       Stoppable stopper,
                       MasterServices master,
                       ServerName serverName,
                       boolean masterRecovery,
                       SplitLogManager.TaskFinisher tf)
                throws InterruptedIOException,
                       org.apache.zookeeper.KeeperException

Its OK to construct this object even when region-servers are not online. It does lookup the orphan tasks in zk but it doesn't block waiting for them to be done.

Parameters:

zkw - the ZK watcher

conf - the HBase configuration

stopper - the stoppable in case anything is wrong

master - the master services

serverName - the master server name

masterRecovery - an indication if the master is in recovery

tf - task finisher

Throws:

org.apache.zookeeper.KeeperException

InterruptedIOException

Method Detail

splitLogDistributed

public long splitLogDistributed(org.apache.hadoop.fs.Path logDir)
                         throws IOException

Parameters:

logDir - one region sever hlog dir path in .logs

Returns:

cumulative size of the logfiles split

Throws:

IOException - if there was an error while splitting any log file

IOException

splitLogDistributed

public long splitLogDistributed(List<org.apache.hadoop.fs.Path> logDirs)
                         throws IOException

The caller will block until all the log files of the given region server have been processed - successfully split or an error is encountered - by an available worker region server. This method must only be called after the region servers have been brought online.

Parameters:

logDirs - List of log dirs to split

Returns:

cumulative size of the logfiles split

Throws:

IOException - If there was an error while splitting any log file

splitLogDistributed

public long splitLogDistributed(Set<ServerName> serverNames,
                                List<org.apache.hadoop.fs.Path> logDirs,
                                org.apache.hadoop.fs.PathFilter filter)
                         throws IOException

The caller will block until all the hbase:meta log files of the given region server have been processed - successfully split or an error is encountered - by an available worker region server. This method must only be called after the region servers have been brought online.

Parameters:

logDirs - List of log dirs to split

filter - the Path filter to select specific files for considering

Returns:

cumulative size of the logfiles split

Throws:

IOException - If there was an error while splitting any log file

deleteRecoveringRegionZNodes

public static void deleteRecoveringRegionZNodes(ZooKeeperWatcher watcher,
                                                List<String> regions)

nodeDataChanged

public void nodeDataChanged(String path)

Overrides:

nodeDataChanged in class ZooKeeperListener

stop

public void stop()

parseLastFlushedSequenceIdFrom

public static long parseLastFlushedSequenceIdFrom(byte[] bytes)

Parameters:

bytes - - Content of a failed region server or recovering region znode.

Returns:

long - The last flushed sequence Id for the region server

isRegionMarkedRecoveringInZK

public static boolean isRegionMarkedRecoveringInZK(ZooKeeperWatcher zkw,
                                                   String regionEncodedName)
                                            throws org.apache.zookeeper.KeeperException

check if /hbase/recovering-regions/ exists. Returns true if exists and set watcher as well.

Parameters:

zkw -

regionEncodedName - region encode name

Returns:

true when /hbase/recovering-regions/ exists

Throws:

org.apache.zookeeper.KeeperException

getRegionFlushedSequenceId

public static org.apache.hadoop.hbase.protobuf.generated.ZooKeeperProtos.RegionStoreSequenceIds getRegionFlushedSequenceId(ZooKeeperWatcher zkw,
                                                                                                                           String serverName,
                                                                                                                           String encodedRegionName)
                                                                                                                    throws IOException

This function is used in distributedLogReplay to fetch last flushed sequence id from ZK

Parameters:

zkw -

serverName -

encodedRegionName -

Returns:

the last flushed sequence ids recorded in ZK of the region for serverName

Throws:

IOException

setRecoveryMode

public void setRecoveryMode(boolean isForInitialization)
                     throws org.apache.zookeeper.KeeperException

This function is to set recovery mode from outstanding split log tasks from before or current configuration setting

Parameters:

isForInitialization -

Throws:

org.apache.zookeeper.KeeperException

InterruptedIOException

getRecoveryMode

public org.apache.hadoop.hbase.protobuf.generated.ZooKeeperProtos.SplitLogTask.RecoveryMode getRecoveryMode()