org.apache.hadoop.hbase.master.locking

类 LockProcedure

java.lang.Object

org.apache.hadoop.hbase.procedure2.Procedure<MasterProcedureEnv>

org.apache.hadoop.hbase.master.locking.LockProcedure

所有已实现的接口:

Comparable<Procedure<MasterProcedureEnv>>, TableProcedureInterface

@InterfaceAudience.Private
public final class LockProcedure
extends Procedure<MasterProcedureEnv>
implements TableProcedureInterface

Procedure to allow blessed clients and external admin tools to take our internal Schema locks used by the procedure framework isolating procedures doing creates/deletes etc. on table/namespace/regions. This procedure when scheduled, acquires specified locks, suspends itself and waits for:

• Call to unlock: if lock request came from the process itself, say master chore.
• Timeout : if lock request came from RPC. On timeout, evaluates if it should continue holding the lock or not based on last heartbeat timestamp.

嵌套类概要

从类继承的嵌套类/接口 org.apache.hadoop.hbase.procedure2.Procedure

Procedure.LockState

从接口继承的嵌套类/接口 org.apache.hadoop.hbase.master.procedure.TableProcedureInterface

TableProcedureInterface.TableOperationType

字段概要

字段

限定符和类型	字段和说明
static int	DEFAULT_LOCAL_MASTER_LOCKS_TIMEOUT_MS
static int	DEFAULT_REMOTE_LOCKS_TIMEOUT_MS
static String	LOCAL_MASTER_LOCKS_TIMEOUT_MS_CONF
static String	REMOTE_LOCKS_TIMEOUT_MS_CONF

从类继承的字段 org.apache.hadoop.hbase.procedure2.Procedure

NO_PROC_ID, NO_TIMEOUT

构造器概要

构造器

构造器和说明
LockProcedure()
LockProcedure(org.apache.hadoop.conf.Configuration conf, RegionInfo[] regionInfos, LockType type, String description, CountDownLatch lockAcquireLatch) Constructor for region lock(s).
LockProcedure(org.apache.hadoop.conf.Configuration conf, String namespace, LockType type, String description, CountDownLatch lockAcquireLatch) Constructor for namespace lock.
LockProcedure(org.apache.hadoop.conf.Configuration conf, TableName tableName, LockType type, String description, CountDownLatch lockAcquireLatch) Constructor for table lock.

方法概要

限定符和类型	方法和说明
protected boolean	abort(MasterProcedureEnv env)
protected Procedure.LockState	acquireLock(MasterProcedureEnv env)
protected void	beforeReplay(MasterProcedureEnv env) On recovery, re-execute from start to acquire the locks.
protected void	deserializeStateData(ProcedureStateSerializer serializer)
protected Procedure<MasterProcedureEnv>[]	execute(MasterProcedureEnv env)
String	getDescription()
TableName	getTableName()
TableProcedureInterface.TableOperationType	getTableOperationType() Given an operation type we can take decisions about what to do with pending operations.
LockType	getType()
boolean	holdLock(MasterProcedureEnv env)
boolean	isLocked()
protected void	releaseLock(MasterProcedureEnv env)
protected void	rollback(MasterProcedureEnv env)
protected void	serializeStateData(ProcedureStateSerializer serializer)
protected boolean	setTimeoutFailure(MasterProcedureEnv env) Re run the procedure after every timeout to write new WAL entries so we don't hold back old WALs.
protected void	toStringClassDetails(StringBuilder builder)
void	unlock(MasterProcedureEnv env)
void	updateHeartBeat() Updates timeout deadline for the lock.

从类继承的方法 org.apache.hadoop.hbase.procedure2.Procedure

addStackIndex, afterReplay, bypass, compareTo, completionCleanup, doExecute, doRollback, elapsedTime, getChildrenLatch, getException, getLastUpdate, getNonceKey, getOwner, getParentProcId, getProcedureMetrics, getProcId, getProcIdHashCode, getProcName, getResult, getRootProcedureId, getRootProcId, getStackIndexes, getState, getSubmittedTime, getTimeout, getTimeoutTimestamp, hasChildren, hasException, hasLock, hasOwner, hasParent, hasTimeout, haveSameParent, incChildrenLatch, isBypass, isFailed, isFinished, isInitializing, isLockedWhenLoading, isRunnable, isSuccess, isWaiting, isYieldAfterExecutionStep, removeStackIndex, setAbortFailure, setChildrenLatch, setFailure, setFailure, setLastUpdate, setNonceKey, setOwner, setOwner, setParentProcId, setProcId, setResult, setRootProcId, setStackIndexes, setState, setSubmittedTime, setTimeout, shouldWaitClientAck, skipPersistence, toString, toStringClass, toStringDetails, toStringSimpleSB, toStringState, updateMetricsOnFinish, updateMetricsOnSubmit, updateTimestamp, waitInitialized, wasExecuted

从类继承的方法 java.lang.Object

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

字段详细资料

DEFAULT_REMOTE_LOCKS_TIMEOUT_MS

public static final int DEFAULT_REMOTE_LOCKS_TIMEOUT_MS

另请参阅:

常量字段值

REMOTE_LOCKS_TIMEOUT_MS_CONF

public static final String REMOTE_LOCKS_TIMEOUT_MS_CONF

另请参阅:

常量字段值

DEFAULT_LOCAL_MASTER_LOCKS_TIMEOUT_MS

public static final int DEFAULT_LOCAL_MASTER_LOCKS_TIMEOUT_MS

另请参阅:

常量字段值

LOCAL_MASTER_LOCKS_TIMEOUT_MS_CONF

public static final String LOCAL_MASTER_LOCKS_TIMEOUT_MS_CONF

另请参阅:

常量字段值

构造器详细资料

LockProcedure

public LockProcedure()

LockProcedure

public LockProcedure(org.apache.hadoop.conf.Configuration conf,
                     String namespace,
                     LockType type,
                     String description,
                     CountDownLatch lockAcquireLatch)
              throws IllegalArgumentException

Constructor for namespace lock.

参数:

lockAcquireLatch - if not null, the latch is decreased when lock is acquired.

抛出:

IllegalArgumentException

LockProcedure

public LockProcedure(org.apache.hadoop.conf.Configuration conf,
                     TableName tableName,
                     LockType type,
                     String description,
                     CountDownLatch lockAcquireLatch)
              throws IllegalArgumentException

Constructor for table lock.

参数:

lockAcquireLatch - if not null, the latch is decreased when lock is acquired.

抛出:

IllegalArgumentException

LockProcedure

public LockProcedure(org.apache.hadoop.conf.Configuration conf,
                     RegionInfo[] regionInfos,
                     LockType type,
                     String description,
                     CountDownLatch lockAcquireLatch)
              throws IllegalArgumentException

Constructor for region lock(s).

参数:

lockAcquireLatch - if not null, the latch is decreased when lock is acquired. Useful for locks acquired locally from master process.

抛出:

IllegalArgumentException - if all regions are not from same table.

方法详细资料

getTableName

public TableName getTableName()

指定者:

getTableName 在接口中 TableProcedureInterface

返回:

the name of the table the procedure is operating on

getTableOperationType

public TableProcedureInterface.TableOperationType getTableOperationType()

从接口复制的说明: TableProcedureInterface

Given an operation type we can take decisions about what to do with pending operations. e.g. if we get a delete and we have some table operation pending (e.g. add column) we can abort those operations.

指定者:

getTableOperationType 在接口中 TableProcedureInterface

返回:

the operation type that the procedure is executing.

updateHeartBeat

public void updateHeartBeat()

Updates timeout deadline for the lock.

setTimeoutFailure

protected boolean setTimeoutFailure(MasterProcedureEnv env)

Re run the procedure after every timeout to write new WAL entries so we don't hold back old WALs.

覆盖:

setTimeoutFailure 在类中 Procedure<MasterProcedureEnv>

返回:

false, so procedure framework doesn't mark this procedure as failure.

unlock

public void unlock(MasterProcedureEnv env)

execute

protected Procedure<MasterProcedureEnv>[] execute(MasterProcedureEnv env)
                                           throws ProcedureSuspendedException

指定者:

execute 在类中 Procedure<MasterProcedureEnv>

抛出:

ProcedureSuspendedException

rollback

protected void rollback(MasterProcedureEnv env)

指定者:

rollback 在类中 Procedure<MasterProcedureEnv>

abort

protected boolean abort(MasterProcedureEnv env)

指定者:

abort 在类中 Procedure<MasterProcedureEnv>

serializeStateData

protected void serializeStateData(ProcedureStateSerializer serializer)
                           throws IOException

指定者:

serializeStateData 在类中 Procedure<MasterProcedureEnv>

抛出:

IOException

deserializeStateData

protected void deserializeStateData(ProcedureStateSerializer serializer)
                             throws IOException

指定者:

deserializeStateData 在类中 Procedure<MasterProcedureEnv>

抛出:

IOException

acquireLock

protected Procedure.LockState acquireLock(MasterProcedureEnv env)

覆盖:

acquireLock 在类中 Procedure<MasterProcedureEnv>

releaseLock

protected void releaseLock(MasterProcedureEnv env)

覆盖:

releaseLock 在类中 Procedure<MasterProcedureEnv>

beforeReplay

protected void beforeReplay(MasterProcedureEnv env)

On recovery, re-execute from start to acquire the locks. Need to explicitly set it to RUNNABLE because the procedure might have been in WAITING_TIMEOUT state when crash happened. In which case, it'll be sent back to timeout queue on recovery, which we don't want since we want to require locks.

覆盖:

beforeReplay 在类中 Procedure<MasterProcedureEnv>

toStringClassDetails

protected void toStringClassDetails(StringBuilder builder)

覆盖:

toStringClassDetails 在类中 Procedure<MasterProcedureEnv>

getType

public LockType getType()

getDescription

public String getDescription()

isLocked

public boolean isLocked()

holdLock

public boolean holdLock(MasterProcedureEnv env)

覆盖:

holdLock 在类中 Procedure<MasterProcedureEnv>