Package org.apache.hadoop.hbase.master.assignment

Class MergeTableRegionsProcedure

java.lang.Object

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

org.apache.hadoop.hbase.procedure2.StateMachineProcedure<MasterProcedureEnv,TState>

org.apache.hadoop.hbase.master.procedure.AbstractStateMachineTableProcedure<org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

org.apache.hadoop.hbase.master.assignment.MergeTableRegionsProcedure

All Implemented Interfaces:

Comparable<Procedure<MasterProcedureEnv>>, TableProcedureInterface

@Private
public class MergeTableRegionsProcedure
extends AbstractStateMachineTableProcedure<org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

The procedure to Merge regions in a table. This procedure takes an exclusive table lock since it is working over multiple regions. It holds the lock for the life of the procedure. Throws exception on construction if determines context hostile to merge (cluster going down or master is shutting down or table is disabled).

Nested Class Summary

Nested classes/interfaces inherited from class org.apache.hadoop.hbase.procedure2.StateMachineProcedure

StateMachineProcedure.Flow

Nested classes/interfaces inherited from class org.apache.hadoop.hbase.procedure2.Procedure

Procedure.LockState

Nested classes/interfaces inherited from interface org.apache.hadoop.hbase.master.procedure.TableProcedureInterface

TableProcedureInterface.TableOperationType

Field Summary

Fields

Modifier and Type	Field	Description
private boolean	force
private static final org.slf4j.Logger	LOG
private RegionInfo	mergedRegion	The resulting merged region.
private ServerName	regionLocation
private RegionInfo[]	regionsToMerge	Two or more regions to merge, the 'merge parents'.

Fields inherited from class org.apache.hadoop.hbase.procedure2.StateMachineProcedure

stateCount

Fields inherited from class org.apache.hadoop.hbase.procedure2.Procedure

NO_PROC_ID, NO_TIMEOUT

Fields inherited from interface org.apache.hadoop.hbase.master.procedure.TableProcedureInterface

DUMMY_NAMESPACE_TABLE_NAME

Constructor Summary

Constructors

Constructor	Description
MergeTableRegionsProcedure()
MergeTableRegionsProcedure(MasterProcedureEnv env, RegionInfo[] regionsToMerge, boolean force)

Method Summary

Modifier and Type	Method	Description
protected boolean	abort(MasterProcedureEnv env)	The abort() call is asynchronous and each procedure must decide how to deal with it, if they want to be abortable.
protected Procedure.LockState	acquireLock(MasterProcedureEnv env)	The user should override this method if they need a lock on an Entity.
private void	checkClosedRegions(MasterProcedureEnv env)
private static void	checkRegionsToMerge(MasterProcedureEnv env, RegionInfo[] regions, boolean force)
private void	cleanupMergedRegion(MasterProcedureEnv env)	Clean up a merged region on rollback after failure.
private TransitRegionStateProcedure[]	createAssignProcedures(MasterProcedureEnv env)
private void	createMergedRegion(MasterProcedureEnv env)	Create merged region.
private static RegionInfo	createMergedRegionInfo(RegionInfo[] regionsToMerge)	Create merged region info by looking at passed in regionsToMerge to figure what extremes for start and end keys to use; merged region needs to have an extent sufficient to cover all regions-to-merge.
private TransitRegionStateProcedure[]	createUnassignProcedures(MasterProcedureEnv env)
protected void	deserializeStateData(ProcedureStateSerializer serializer)	Called on store load to allow the user to decode the previously serialized state.
protected StateMachineProcedure.Flow	executeFromState(MasterProcedureEnv env, org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState state)	called to perform a single step of the specified 'state' of the procedure
protected org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState	getInitialState()	Return the initial state object that will be used for the first call to executeFromState().
(package private) RegionInfo	getMergedRegion()	Returns The merged region.
protected ProcedureMetrics	getProcedureMetrics(MasterProcedureEnv env)	Override this method to provide procedure specific counters for submitted count, failed count and time histogram.
private int	getRegionReplication(MasterProcedureEnv env)
private ServerName	getServerName(MasterProcedureEnv env)	The procedure could be restarted from a different machine.
protected org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState	getState(int stateId)	Convert an ordinal (or state id) to an Enum (or more descriptive) state object.
protected int	getStateId(org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState state)	Convert the Enum (or more descriptive) state object to an ordinal (or state id).
TableName	getTableName()	Returns the name of the table the procedure is operating on
TableProcedureInterface.TableOperationType	getTableOperationType()	Given an operation type we can take decisions about what to do with pending operations.
protected boolean	holdLock(MasterProcedureEnv env)	Used to keep the procedure lock even when the procedure is yielding or suspended.
private boolean	isMergeable(MasterProcedureEnv env, RegionState rs)
protected boolean	isRollbackSupported(org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState state)	Used by the default implementation of abort() to know if the current state can be aborted and rollback can be triggered.
private List<org.apache.hadoop.fs.Path>	mergeStoreFiles(MasterProcedureEnv env, HRegionFileSystem regionFs, HRegionFileSystem mergeRegionFs, RegionInfo mergedRegion)
private void	postCompletedMergeRegions(MasterProcedureEnv env)	Post merge region action
private void	postMergeRegionsCommit(MasterProcedureEnv env)	Post merge region action
private void	postRollBackMergeRegions(MasterProcedureEnv env)	Action after rollback a merge table regions action.
private void	preMergeRegions(MasterProcedureEnv env)	Pre merge region action
private void	preMergeRegionsCommit(MasterProcedureEnv env)	Post merge region action
private boolean	prepareMergeRegion(MasterProcedureEnv env)	Prepare merge and do some check
protected void	releaseLock(MasterProcedureEnv env)	The user should override this method, and release lock if necessary.
private void	removeNonDefaultReplicas(MasterProcedureEnv env)
private void	rollbackCloseRegionsForMerge(MasterProcedureEnv env)	Rollback close regions
private void	rollbackPrepareMerge(MasterProcedureEnv env)	Action for rollback a merge table after prepare merge
protected void	rollbackState(MasterProcedureEnv env, org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState state)	To rollback MergeTableRegionsProcedure, two AssignProcedures are asynchronously submitted for each region to be merged (rollback doesn't wait on the completion of the AssignProcedures) .
protected void	serializeStateData(ProcedureStateSerializer serializer)	The user-level code of the procedure may have some state to persist (e.g.
private void	setRegionStateToMerging(MasterProcedureEnv env)	Set the region states to MERGING state
void	toStringClassDetails(StringBuilder sb)	Extend the toString() information with the procedure details e.g.
private void	updateMetaForMergedRegions(MasterProcedureEnv env)	Add merged region to META and delete original regions.
private void	writeMaxSequenceIdFile(MasterProcedureEnv env)

Methods inherited from class org.apache.hadoop.hbase.master.procedure.AbstractStateMachineTableProcedure

checkOnline, checkTableModifiable, getUser, getWALRegionDir, isTableEnabled, preflightChecks, releaseSyncLatch, setUser, waitInitialized

Methods inherited from class org.apache.hadoop.hbase.procedure2.StateMachineProcedure

addChildProcedure, execute, failIfAborted, getCurrentState, getCurrentStateId, getCycles, isEofState, isRollbackSupported, isYieldAfterExecutionStep, isYieldBeforeExecuteFromState, rollback, setNextState, toStringState

Methods inherited from class org.apache.hadoop.hbase.procedure2.Procedure

addStackIndex, afterReplay, beforeReplay, bypass, compareTo, completionCleanup, doExecute, doRollback, elapsedTime, getChildrenLatch, getException, getLastUpdate, getNonceKey, getOwner, getParentProcId, 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, removeStackIndex, setAbortFailure, setChildrenLatch, setExecuted, setFailure, setFailure, setLastUpdate, setNonceKey, setOwner, setOwner, setParentProcId, setProcId, setResult, setRootProcId, setStackIndexes, setState, setSubmittedTime, setTimeout, setTimeoutFailure, shouldWaitClientAck, skipPersistence, suspend, toString, toStringClass, toStringDetails, toStringSimpleSB, updateMetricsOnFinish, updateMetricsOnSubmit, updateTimestamp, wasExecuted

Methods inherited from class java.lang.Object

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

Field Details

LOG

private static final org.slf4j.Logger LOG

regionLocation

private ServerName regionLocation

regionsToMerge

private RegionInfo[] regionsToMerge

Two or more regions to merge, the 'merge parents'.

mergedRegion

private RegionInfo mergedRegion

The resulting merged region.

force

private boolean force

Constructor Details

MergeTableRegionsProcedure

public MergeTableRegionsProcedure()

MergeTableRegionsProcedure

public MergeTableRegionsProcedure(MasterProcedureEnv env,
 RegionInfo[] regionsToMerge,
 boolean force)
                           throws IOException

Throws:

IOException

Method Details

checkRegionsToMerge

private static void checkRegionsToMerge(MasterProcedureEnv env,
 RegionInfo[] regions,
 boolean force)
                                 throws MergeRegionException

Throws:

MergeRegionException - If unable to merge regions for whatever reasons.

createMergedRegionInfo

private static RegionInfo createMergedRegionInfo(RegionInfo[] regionsToMerge)

Create merged region info by looking at passed in regionsToMerge to figure what extremes for start and end keys to use; merged region needs to have an extent sufficient to cover all regions-to-merge.

executeFromState

protected StateMachineProcedure.Flow executeFromState(MasterProcedureEnv env,
 org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState state)

Description copied from class: StateMachineProcedure

called to perform a single step of the specified 'state' of the procedure

Specified by:

executeFromState in class StateMachineProcedure<MasterProcedureEnv,org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

state - state to execute

Returns:

Flow.NO_MORE_STATE if the procedure is completed, Flow.HAS_MORE_STATE if there is another step.

rollbackState

protected void rollbackState(MasterProcedureEnv env,
 org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState state)
                      throws IOException

To rollback MergeTableRegionsProcedure, two AssignProcedures are asynchronously submitted for each region to be merged (rollback doesn't wait on the completion of the AssignProcedures) . This can be improved by changing rollback() to support sub-procedures. See HBASE-19851 for details.

Specified by:

rollbackState in class StateMachineProcedure<MasterProcedureEnv,org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

state - state to rollback

Throws:

IOException - temporary failure, the rollback will retry later

isRollbackSupported

protected boolean isRollbackSupported(org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState state)

Description copied from class: StateMachineProcedure

Used by the default implementation of abort() to know if the current state can be aborted and rollback can be triggered.

Overrides:

isRollbackSupported in class StateMachineProcedure<MasterProcedureEnv,org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

removeNonDefaultReplicas

private void removeNonDefaultReplicas(MasterProcedureEnv env)
                               throws IOException

Throws:

IOException

checkClosedRegions

private void checkClosedRegions(MasterProcedureEnv env)
                         throws IOException

Throws:

IOException

getState

protected org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState getState(int stateId)

Description copied from class: StateMachineProcedure

Convert an ordinal (or state id) to an Enum (or more descriptive) state object.

Specified by:

getState in class StateMachineProcedure<MasterProcedureEnv,org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

Parameters:

stateId - the ordinal() of the state enum (or state id)

Returns:

the state enum object

getStateId

protected int getStateId(org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState state)

Description copied from class: StateMachineProcedure

Convert the Enum (or more descriptive) state object to an ordinal (or state id).

Specified by:

getStateId in class StateMachineProcedure<MasterProcedureEnv,org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

Parameters:

state - the state enum object

Returns:

stateId the ordinal() of the state enum (or state id)

getInitialState

protected org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState getInitialState()

Description copied from class: StateMachineProcedure

Return the initial state object that will be used for the first call to executeFromState().

Specified by:

getInitialState in class StateMachineProcedure<MasterProcedureEnv,org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

Returns:

the initial state enum object

serializeStateData

protected void serializeStateData(ProcedureStateSerializer serializer)
                           throws IOException

Description copied from class: Procedure

The user-level code of the procedure may have some state to persist (e.g. input arguments or current position in the processing state) to be able to resume on failure.

Overrides:

serializeStateData in class StateMachineProcedure<MasterProcedureEnv,org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

Parameters:

serializer - stores the serializable state

Throws:

IOException

deserializeStateData

protected void deserializeStateData(ProcedureStateSerializer serializer)
                             throws IOException

Description copied from class: Procedure

Called on store load to allow the user to decode the previously serialized state.

Overrides:

deserializeStateData in class StateMachineProcedure<MasterProcedureEnv,org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

Parameters:

serializer - contains the serialized state

Throws:

IOException

toStringClassDetails

public void toStringClassDetails(StringBuilder sb)

Description copied from class: Procedure

Extend the toString() information with the procedure details e.g. className and parameters

Overrides:

toStringClassDetails in class AbstractStateMachineTableProcedure<org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

Parameters:

sb - the string builder to use to append the proc specific information

acquireLock

protected Procedure.LockState acquireLock(MasterProcedureEnv env)

Description copied from class: Procedure

The user should override this method if they need a lock on an Entity. A lock can be anything, and it is up to the implementor. The Procedure Framework will call this method just before it invokes Procedure.execute(Object). It calls Procedure.releaseLock(Object) after the call to execute.

If you need to hold the lock for the life of the Procedure -- i.e. you do not want any other Procedure interfering while this Procedure is running, see Procedure.holdLock(Object).

Example: in our Master we can execute request in parallel for different tables. We can create t1 and create t2 and these creates can be executed at the same time. Anything else on t1/t2 is queued waiting that specific table create to happen.

There are 3 LockState:

• LOCK_ACQUIRED should be returned when the proc has the lock and the proc is ready to execute.
• LOCK_YIELD_WAIT should be returned when the proc has not the lock and the framework should take care of readding the procedure back to the runnable set for retry
• LOCK_EVENT_WAIT should be returned when the proc has not the lock and someone will take care of readding the procedure back to the runnable set when the lock is available.

Overrides:

acquireLock in class AbstractStateMachineTableProcedure<org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

Returns:

the lock state as described above.

releaseLock

protected void releaseLock(MasterProcedureEnv env)

Description copied from class: Procedure

The user should override this method, and release lock if necessary.

Overrides:

releaseLock in class AbstractStateMachineTableProcedure<org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

holdLock

protected boolean holdLock(MasterProcedureEnv env)

Description copied from class: Procedure

Used to keep the procedure lock even when the procedure is yielding or suspended.

Overrides:

holdLock in class Procedure<MasterProcedureEnv>

Returns:

true if the procedure should hold on the lock until completionCleanup()

getTableName

public TableName getTableName()

Description copied from interface: TableProcedureInterface

Returns the name of the table the procedure is operating on

Specified by:

getTableName in interface TableProcedureInterface

Specified by:

getTableName in class AbstractStateMachineTableProcedure<org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

getTableOperationType

public TableProcedureInterface.TableOperationType getTableOperationType()

Description copied from interface: 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.

Specified by:

getTableOperationType in interface TableProcedureInterface

Specified by:

getTableOperationType in class AbstractStateMachineTableProcedure<org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>

Returns:

the operation type that the procedure is executing.

getProcedureMetrics

protected ProcedureMetrics getProcedureMetrics(MasterProcedureEnv env)

Description copied from class: Procedure

Override this method to provide procedure specific counters for submitted count, failed count and time histogram.

Overrides:

getProcedureMetrics in class Procedure<MasterProcedureEnv>

Parameters:

env - The environment passed to the procedure executor

Returns:

Container object for procedure related metric

prepareMergeRegion

private boolean prepareMergeRegion(MasterProcedureEnv env)
                            throws IOException

Prepare merge and do some check

Throws:

IOException

isMergeable

private boolean isMergeable(MasterProcedureEnv env,
 RegionState rs)
                     throws IOException

Throws:

IOException

rollbackPrepareMerge

private void rollbackPrepareMerge(MasterProcedureEnv env)
                           throws IOException

Action for rollback a merge table after prepare merge

Throws:

IOException

preMergeRegions

private void preMergeRegions(MasterProcedureEnv env)
                      throws IOException

Pre merge region action

Parameters:

env - MasterProcedureEnv

Throws:

IOException

postRollBackMergeRegions

private void postRollBackMergeRegions(MasterProcedureEnv env)
                               throws IOException

Action after rollback a merge table regions action.

Throws:

IOException

setRegionStateToMerging

private void setRegionStateToMerging(MasterProcedureEnv env)

Set the region states to MERGING state

createMergedRegion

private void createMergedRegion(MasterProcedureEnv env)
                         throws IOException

Create merged region. The way the merge works is that we make a 'merges' temporary directory in the FIRST parent region to merge (Do not change this without also changing the rollback where we look in this FIRST region for the merge dir). We then collect here references to all the store files in all the parent regions including those of the FIRST parent region into a subdirectory, named for the resultant merged region. We then call commitMergeRegion. It finds this subdirectory of storefile references and moves them under the new merge region (creating the region layout as side effect). After assign of the new merge region, we will run a compaction. This will undo the references but the reference files remain in place until the archiver runs (which it does on a period as a chore in the RegionServer that hosts the merge region -- see CompactedHFilesDischarger). Once the archiver has moved aside the no-longer used references, the merge region no longer has references. The catalog janitor will notice when it runs next and it will remove the old parent regions.

Throws:

IOException

mergeStoreFiles

private List<org.apache.hadoop.fs.Path> mergeStoreFiles(MasterProcedureEnv env,
 HRegionFileSystem regionFs,
 HRegionFileSystem mergeRegionFs,
 RegionInfo mergedRegion)
                                                 throws IOException

Throws:

IOException

cleanupMergedRegion

private void cleanupMergedRegion(MasterProcedureEnv env)
                          throws IOException

Clean up a merged region on rollback after failure.

Throws:

IOException

rollbackCloseRegionsForMerge

private void rollbackCloseRegionsForMerge(MasterProcedureEnv env)
                                   throws IOException

Rollback close regions

Throws:

IOException

createUnassignProcedures

private TransitRegionStateProcedure[] createUnassignProcedures(MasterProcedureEnv env)
                                                        throws IOException

Throws:

IOException

createAssignProcedures

private TransitRegionStateProcedure[] createAssignProcedures(MasterProcedureEnv env)
                                                      throws IOException

Throws:

IOException

getRegionReplication

private int getRegionReplication(MasterProcedureEnv env)
                          throws IOException

Throws:

IOException

preMergeRegionsCommit

private void preMergeRegionsCommit(MasterProcedureEnv env)
                            throws IOException

Post merge region action

Parameters:

env - MasterProcedureEnv

Throws:

IOException

updateMetaForMergedRegions

private void updateMetaForMergedRegions(MasterProcedureEnv env)
                                 throws IOException

Add merged region to META and delete original regions.

Throws:

IOException

postMergeRegionsCommit

private void postMergeRegionsCommit(MasterProcedureEnv env)
                             throws IOException

Post merge region action

Parameters:

env - MasterProcedureEnv

Throws:

IOException

postCompletedMergeRegions

private void postCompletedMergeRegions(MasterProcedureEnv env)
                                throws IOException

Post merge region action

Parameters:

env - MasterProcedureEnv

Throws:

IOException

getServerName

private ServerName getServerName(MasterProcedureEnv env)

The procedure could be restarted from a different machine. If the variable is null, we need to retrieve it.

Parameters:

env - MasterProcedureEnv

writeMaxSequenceIdFile

private void writeMaxSequenceIdFile(MasterProcedureEnv env)
                             throws IOException

Throws:

IOException

getMergedRegion

RegionInfo getMergedRegion()

Returns The merged region. Maybe be null if called to early or we failed.

abort

protected boolean abort(MasterProcedureEnv env)

Description copied from class: Procedure

The abort() call is asynchronous and each procedure must decide how to deal with it, if they want to be abortable. The simplest implementation is to have an AtomicBoolean set in the abort() method and then the execute() will check if the abort flag is set or not. abort() may be called multiple times from the client, so the implementation must be idempotent.

NOTE: abort() is not like Thread.interrupt(). It is just a notification that allows the procedure implementor abort.

Overrides:

abort in class StateMachineProcedure<MasterProcedureEnv,org.apache.hadoop.hbase.shaded.protobuf.generated.MasterProcedureProtos.MergeTableRegionsState>