org.rocksdb

Class TransactionDBOptions

java.lang.Object

org.rocksdb.AbstractNativeReference

org.rocksdb.AbstractImmutableNativeReference

org.rocksdb.RocksObject

org.rocksdb.TransactionDBOptions

All Implemented Interfaces:

java.lang.AutoCloseable

public class TransactionDBOptions
extends RocksObject

Field Summary

Fields inherited from class org.rocksdb.RocksObject

nativeHandle_

Fields inherited from class org.rocksdb.AbstractImmutableNativeReference

owningHandle_

Constructor Summary

Constructors

Constructor and Description
TransactionDBOptions()

Method Summary

Modifier and Type	Method and Description
protected void	disposeInternal(long handle)
long	getDefaultLockTimeout() The wait timeout in milliseconds when writing a key OUTSIDE of a transaction (ie by calling RocksDB.put(byte[], byte[]), RocksDB.merge(byte[], byte[]), RocksDB.delete(byte[]) or RocksDB.write(org.rocksdb.WriteOptions, org.rocksdb.WriteBatch) directly).
long	getMaxNumLocks() Specifies the maximum number of keys that can be locked at the same time per column family.
long	getNumStripes() The number of sub-tables per lock table (per column family)
long	getTransactionLockTimeout() The default wait timeout in milliseconds when a transaction attempts to lock a key if not specified by TransactionOptions.setLockTimeout(long) If 0, no waiting is done if a lock cannot instantly be acquired.
TxnDBWritePolicy	getWritePolicy() The policy for when to write the data into the DB.
TransactionDBOptions	setDefaultLockTimeout(long defaultLockTimeout) If positive, specifies the wait timeout in milliseconds when writing a key OUTSIDE of a transaction (ie by calling RocksDB.put(byte[], byte[]), RocksDB.merge(byte[], byte[]), RocksDB.delete(byte[]) or RocksDB.write(org.rocksdb.WriteOptions, org.rocksdb.WriteBatch) directly).
TransactionDBOptions	setMaxNumLocks(long maxNumLocks) Specifies the maximum number of keys that can be locked at the same time per column family.
TransactionDBOptions	setNumStripes(long numStripes) Increasing this value will increase the concurrency by dividing the lock table (per column family) into more sub-tables, each with their own separate mutex.
TransactionDBOptions	setTransactionLockTimeout(long transactionLockTimeout) If positive, specifies the default wait timeout in milliseconds when a transaction attempts to lock a key if not specified by TransactionOptions.setLockTimeout(long) If 0, no waiting is done if a lock cannot instantly be acquired.
TransactionDBOptions	setWritePolicy(TxnDBWritePolicy writePolicy) The policy for when to write the data into the DB.

Methods inherited from class org.rocksdb.RocksObject

disposeInternal, getNativeHandle

Methods inherited from class org.rocksdb.AbstractImmutableNativeReference

close, disOwnNativeHandle, isOwningHandle

Methods inherited from class java.lang.Object

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

Constructor Detail

TransactionDBOptions

public TransactionDBOptions()

Method Detail

getMaxNumLocks

public long getMaxNumLocks()

Specifies the maximum number of keys that can be locked at the same time per column family.

If the number of locked keys is greater than #getMaxNumLocks(), transaction writes (or GetForUpdate) will return an error.

Returns:

The maximum number of keys that can be locked

setMaxNumLocks

public TransactionDBOptions setMaxNumLocks(long maxNumLocks)

Specifies the maximum number of keys that can be locked at the same time per column family.

If the number of locked keys is greater than getMaxNumLocks(), transaction writes (or GetForUpdate) will return an error.

Parameters:

maxNumLocks - The maximum number of keys that can be locked; If this value is not positive, no limit will be enforced.

Returns:

this TransactionDBOptions instance

getNumStripes

public long getNumStripes()

The number of sub-tables per lock table (per column family)

Returns:

The number of sub-tables

setNumStripes

public TransactionDBOptions setNumStripes(long numStripes)

Increasing this value will increase the concurrency by dividing the lock table (per column family) into more sub-tables, each with their own separate mutex.

Default: 16

Parameters:

numStripes - The number of sub-tables

Returns:

this TransactionDBOptions instance

getTransactionLockTimeout

public long getTransactionLockTimeout()

The default wait timeout in milliseconds when a transaction attempts to lock a key if not specified by TransactionOptions.setLockTimeout(long) If 0, no waiting is done if a lock cannot instantly be acquired. If negative, there is no timeout.

Returns:

the default wait timeout in milliseconds

setTransactionLockTimeout

public TransactionDBOptions setTransactionLockTimeout(long transactionLockTimeout)

If positive, specifies the default wait timeout in milliseconds when a transaction attempts to lock a key if not specified by TransactionOptions.setLockTimeout(long) If 0, no waiting is done if a lock cannot instantly be acquired. If negative, there is no timeout. Not using a timeout is not recommended as it can lead to deadlocks. Currently, there is no deadlock-detection to recover from a deadlock.

Default: 1000

Parameters:

transactionLockTimeout - the default wait timeout in milliseconds

Returns:

this TransactionDBOptions instance

getDefaultLockTimeout

public long getDefaultLockTimeout()

The wait timeout in milliseconds when writing a key OUTSIDE of a transaction (ie by calling RocksDB.put(byte[], byte[]), RocksDB.merge(byte[], byte[]), RocksDB.delete(byte[]) or RocksDB.write(org.rocksdb.WriteOptions, org.rocksdb.WriteBatch) directly).

If 0, no waiting is done if a lock cannot instantly be acquired. If negative, there is no timeout and will block indefinitely when acquiring a lock.

Returns:

the timeout in milliseconds when writing a key OUTSIDE of a transaction

setDefaultLockTimeout

public TransactionDBOptions setDefaultLockTimeout(long defaultLockTimeout)

If positive, specifies the wait timeout in milliseconds when writing a key OUTSIDE of a transaction (ie by calling RocksDB.put(byte[], byte[]), RocksDB.merge(byte[], byte[]), RocksDB.delete(byte[]) or RocksDB.write(org.rocksdb.WriteOptions, org.rocksdb.WriteBatch) directly).

If 0, no waiting is done if a lock cannot instantly be acquired. If negative, there is no timeout and will block indefinitely when acquiring a lock.

Not using a timeout can lead to deadlocks. Currently, there is no deadlock-detection to recover from a deadlock. While DB writes cannot deadlock with other DB writes, they can deadlock with a transaction. A negative timeout should only be used if all transactions have a small expiration set.

Default: 1000

Parameters:

defaultLockTimeout - the timeout in milliseconds when writing a key OUTSIDE of a transaction

Returns:

this TransactionDBOptions instance

getWritePolicy

public TxnDBWritePolicy getWritePolicy()

The policy for when to write the data into the DB. The default policy is to write only the committed data TxnDBWritePolicy.WRITE_COMMITTED. The data could be written before the commit phase. The DB then needs to provide the mechanisms to tell apart committed from uncommitted data.

Returns:

The write policy.

setWritePolicy

public TransactionDBOptions setWritePolicy(TxnDBWritePolicy writePolicy)

The policy for when to write the data into the DB. The default policy is to write only the committed data TxnDBWritePolicy.WRITE_COMMITTED. The data could be written before the commit phase. The DB then needs to provide the mechanisms to tell apart committed from uncommitted data.

Parameters:

writePolicy - The write policy.

Returns:

this TransactionDBOptions instance

disposeInternal

protected final void disposeInternal(long handle)

Specified by:

disposeInternal in class RocksObject