學習自https://www.cnblogs.com/yeyang/p/10226284.html
百度uid-generator源碼
https://github.com/baidu/uid-generator
解析uid-generator
uid-generator是基於Twitter開源的snowflake算法實現的。snowflake雪花算法上文有介紹 傳送門
snowflake將long的64位分爲了3部分,時間戳、工作機器id和序列號、其中,時間戳部分的時間單位一般爲毫秒。也就是說1臺工作機器1毫秒可產生4096個id(2的12次方)。
源碼實現分析
與原始的snowflake算法不同,uid-generator支持自定義時間戳、工作機器id和序列號等各部分的位數,以應用於不同場景。默認分配方式如下。
-
sign(1bit)
固定1bit符號標識,即生成的UID爲正數。 -
delta seconds (28 bits)
當前時間,相對於時間基點"2016-05-20"的增量值,單位:秒,最多可支持約8.7年(注意:1. 這裏的單位是秒,而不是毫秒! 2.注意這裏的用詞,是“最多”可支持8.7年,爲什麼是“最多”,後面會講) -
worker id (22 bits)
機器id,最多可支持約420w次機器啓動。內置實現爲在啓動時由數據庫分配,默認分配策略爲用後即棄,後續可提供複用策略。 -
sequence (13 bits)
每秒下的併發序列,13 bits可支持每秒8192個併發。(注意下這個地方,默認支持qps最大爲8192個)
DefaultUidGenerator
DefaultUidGenerator的產生id的方法與基本上就是常見的snowflake算法實現,僅有一些不同,如以秒爲爲單位而不是毫秒。
DefaultUidGenerator的產生id的方法如下。
/*
* Copyright (c) 2017 Baidu, Inc. All Rights Reserve.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package com.baidu.fsg.uid.impl;
import java.util.Date;
import java.util.concurrent.TimeUnit;
import org.apache.commons.lang.StringUtils;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.beans.factory.InitializingBean;
import com.baidu.fsg.uid.BitsAllocator;
import com.baidu.fsg.uid.UidGenerator;
import com.baidu.fsg.uid.exception.UidGenerateException;
import com.baidu.fsg.uid.utils.DateUtils;
import com.baidu.fsg.uid.worker.WorkerIdAssigner;
/**
* Represents an implementation of {@link UidGenerator}
*
* The unique id has 64bits (long), default allocated as blow:<br>
* <li>sign: The highest bit is 0
* <li>delta seconds: The next 28 bits, represents delta seconds since a customer epoch(2016-05-20 00:00:00.000).
* Supports about 8.7 years until to 2024-11-20 21:24:16
* <li>worker id: The next 22 bits, represents the worker's id which assigns based on database, max id is about 420W
* <li>sequence: The next 13 bits, represents a sequence within the same second, max for 8192/s<br><br>
*
* The {@link DefaultUidGenerator#parseUID(long)} is a tool method to parse the bits
*
* <pre>{@code
* +------+----------------------+----------------+-----------+
* | sign | delta seconds | worker node id | sequence |
* +------+----------------------+----------------+-----------+
* 1bit 28bits 22bits 13bits
* }</pre>
*
* You can also specified the bits by Spring property setting.
* <li>timeBits: default as 28
* <li>workerBits: default as 22
* <li>seqBits: default as 13
* <li>epochStr: Epoch date string format 'yyyy-MM-dd'. Default as '2016-05-20'<p>
*
* <b>Note that:</b> The total bits must be 64 -1
*
* @author yutianbao
*/
public class DefaultUidGenerator implements UidGenerator, InitializingBean {
private static final Logger LOGGER = LoggerFactory.getLogger(DefaultUidGenerator.class);
/** Bits allocate */
protected int timeBits = 28;
protected int workerBits = 22;
protected int seqBits = 13;
/** Customer epoch, unit as second. For example 2016-05-20 (ms: 1463673600000)*/
protected String epochStr = "2016-05-20";
protected long epochSeconds = TimeUnit.MILLISECONDS.toSeconds(1463673600000L);
/** Stable fields after spring bean initializing */
protected BitsAllocator bitsAllocator;
protected long workerId;
/** Volatile fields caused by nextId() */
protected long sequence = 0L;
protected long lastSecond = -1L;
/** Spring property */
protected WorkerIdAssigner workerIdAssigner;
@Override
public void afterPropertiesSet() throws Exception {
// initialize bits allocator
bitsAllocator = new BitsAllocator(timeBits, workerBits, seqBits);
// initialize worker id
workerId = workerIdAssigner.assignWorkerId();
if (workerId > bitsAllocator.getMaxWorkerId()) {
throw new RuntimeException("Worker id " + workerId + " exceeds the max " + bitsAllocator.getMaxWorkerId());
}
LOGGER.info("Initialized bits(1, {}, {}, {}) for workerID:{}", timeBits, workerBits, seqBits, workerId);
}
@Override
public long getUID() throws UidGenerateException {
try {
return nextId();
} catch (Exception e) {
LOGGER.error("Generate unique id exception. ", e);
throw new UidGenerateException(e);
}
}
@Override
public String parseUID(long uid) {
long totalBits = BitsAllocator.TOTAL_BITS;
long signBits = bitsAllocator.getSignBits();
long timestampBits = bitsAllocator.getTimestampBits();
long workerIdBits = bitsAllocator.getWorkerIdBits();
long sequenceBits = bitsAllocator.getSequenceBits();
// parse UID
long sequence = (uid << (totalBits - sequenceBits)) >>> (totalBits - sequenceBits);
long workerId = (uid << (timestampBits + signBits)) >>> (totalBits - workerIdBits);
long deltaSeconds = uid >>> (workerIdBits + sequenceBits);
Date thatTime = new Date(TimeUnit.SECONDS.toMillis(epochSeconds + deltaSeconds));
String thatTimeStr = DateUtils.formatByDateTimePattern(thatTime);
// format as string
return String.format("{\"UID\":\"%d\",\"timestamp\":\"%s\",\"workerId\":\"%d\",\"sequence\":\"%d\"}",
uid, thatTimeStr, workerId, sequence);
}
/**
* Get UID
*
* @return UID
* @throws UidGenerateException in the case: Clock moved backwards; Exceeds the max timestamp
*/
protected synchronized long nextId() {
long currentSecond = getCurrentSecond();
// Clock moved backwards, refuse to generate uid
if (currentSecond < lastSecond) {
long refusedSeconds = lastSecond - currentSecond;
throw new UidGenerateException("Clock moved backwards. Refusing for %d seconds", refusedSeconds);
}
// At the same second, increase sequence
if (currentSecond == lastSecond) {
sequence = (sequence + 1) & bitsAllocator.getMaxSequence();
// Exceed the max sequence, we wait the next second to generate uid
if (sequence == 0) {
currentSecond = getNextSecond(lastSecond);
}
// At the different second, sequence restart from zero
} else {
sequence = 0L;
}
lastSecond = currentSecond;
// Allocate bits for UID
return bitsAllocator.allocate(currentSecond - epochSeconds, workerId, sequence);
}
/**
* Get next millisecond
*/
private long getNextSecond(long lastTimestamp) {
long timestamp = getCurrentSecond();
while (timestamp <= lastTimestamp) {
timestamp = getCurrentSecond();
}
return timestamp;
}
/**
* Get current second
*/
private long getCurrentSecond() {
long currentSecond = TimeUnit.MILLISECONDS.toSeconds(System.currentTimeMillis());
if (currentSecond - epochSeconds > bitsAllocator.getMaxDeltaSeconds()) {
throw new UidGenerateException("Timestamp bits is exhausted. Refusing UID generate. Now: " + currentSecond);
}
return currentSecond;
}
/**
* Setters for spring property
*/
public void setWorkerIdAssigner(WorkerIdAssigner workerIdAssigner) {
this.workerIdAssigner = workerIdAssigner;
}
public void setTimeBits(int timeBits) {
if (timeBits > 0) {
this.timeBits = timeBits;
}
}
public void setWorkerBits(int workerBits) {
if (workerBits > 0) {
this.workerBits = workerBits;
}
}
public void setSeqBits(int seqBits) {
if (seqBits > 0) {
this.seqBits = seqBits;
}
}
public void setEpochStr(String epochStr) {
if (StringUtils.isNotBlank(epochStr)) {
this.epochStr = epochStr;
this.epochSeconds = TimeUnit.MILLISECONDS.toSeconds(DateUtils.parseByDayPattern(epochStr).getTime());
}
}
}
核心代碼:
/**
* Get UID
*
* @return UID
* @throws UidGenerateException in the case: Clock moved backwards; Exceeds the max timestamp
*/
protected synchronized long nextId() {
long currentSecond = getCurrentSecond();
// Clock moved backwards, refuse to generate uid
if (currentSecond < lastSecond) {
long refusedSeconds = lastSecond - currentSecond;
throw new UidGenerateException("Clock moved backwards. Refusing for %d seconds", refusedSeconds);
}
// At the same second, increase sequence
if (currentSecond == lastSecond) {
sequence = (sequence + 1) & bitsAllocator.getMaxSequence();
// Exceed the max sequence, we wait the next second to generate uid
if (sequence == 0) {
currentSecond = getNextSecond(lastSecond);
}
// At the different second, sequence restart from zero
} else {
sequence = 0L;
}
lastSecond = currentSecond;
// Allocate bits for UID
return bitsAllocator.allocate(currentSecond - epochSeconds, workerId, sequence);
}
CachedUidGenerator
CachedUidGenerator支持緩存生成的id。
基本實現原理
關於CachedUidGenerator,文檔上是這樣介紹的。
在實現上, UidGenerator通過借用未來時間來解決sequence天然存在的併發限制; 採用RingBuffer來緩存已生成的UID, 並行化UID的生產和消費, 同時對CacheLine補齊,避免了由RingBuffer帶來的硬件級「僞共享」問題. 最終單機QPS可達600萬。
【採用RingBuffer來緩存已生成的UID, 並行化UID的生產和消費】
使用RingBuffer緩存生成的id。RingBuffer是個環形數組,默認大小爲8192個,裏面緩存着生成的id。
獲取id
會從ringbuffer中拿一個id,支持併發獲取
填充id
RingBuffer填充時機
-
程序啓動時,將RingBuffer填充滿,緩存着8192個id
-
在調用getUID()獲取id時,檢測到RingBuffer中的剩餘id個數小於總個數的50%,將RingBuffer填充滿,使其緩存8192個id
-
定時填充(可配置是否使用以及定時任務的週期)
【UidGenerator通過借用未來時間來解決sequence天然存在的併發限制】
因爲delta seconds部分是以秒爲單位的,所以1個worker 1秒內最多生成的id書爲8192個(2的13次方)。
從上可知,支持的最大qps爲8192,所以通過緩存id來提高吞吐量。
爲什麼叫藉助未來時間?
因爲每秒最多生成8192個id,當1秒獲取id數多於8192時,RingBuffer中的id很快消耗完畢,在填充RingBuffer時,生成的id的delta seconds 部分只能使用未來的時間。
(因爲使用了未來的時間來生成id,所以上面說的是,【最多】可支持約8.7年)
/*
* Copyright (c) 2017 Baidu, Inc. All Rights Reserve.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package com.baidu.fsg.uid.impl;
import java.util.ArrayList;
import java.util.List;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.beans.factory.DisposableBean;
import org.springframework.util.Assert;
import com.baidu.fsg.uid.BitsAllocator;
import com.baidu.fsg.uid.UidGenerator;
import com.baidu.fsg.uid.buffer.BufferPaddingExecutor;
import com.baidu.fsg.uid.buffer.RejectedPutBufferHandler;
import com.baidu.fsg.uid.buffer.RejectedTakeBufferHandler;
import com.baidu.fsg.uid.buffer.RingBuffer;
import com.baidu.fsg.uid.exception.UidGenerateException;
/**
* Represents a cached implementation of {@link UidGenerator} extends
* from {@link DefaultUidGenerator}, based on a lock free {@link RingBuffer}<p>
*
* The spring properties you can specified as below:<br>
* <li><b>boostPower:</b> RingBuffer size boost for a power of 2, Sample: boostPower is 3, it means the buffer size
* will be <code>({@link BitsAllocator#getMaxSequence()} + 1) <<
* {@link #boostPower}</code>, Default as {@value #DEFAULT_BOOST_POWER}
* <li><b>paddingFactor:</b> Represents a percent value of (0 - 100). When the count of rest available UIDs reach the
* threshold, it will trigger padding buffer. Default as{@link RingBuffer#DEFAULT_PADDING_PERCENT}
* Sample: paddingFactor=20, bufferSize=1000 -> threshold=1000 * 20 /100, padding buffer will be triggered when tail-cursor<threshold
* <li><b>scheduleInterval:</b> Padding buffer in a schedule, specify padding buffer interval, Unit as second
* <li><b>rejectedPutBufferHandler:</b> Policy for rejected put buffer. Default as discard put request, just do logging
* <li><b>rejectedTakeBufferHandler:</b> Policy for rejected take buffer. Default as throwing up an exception
*
* @author yutianbao
*/
public class CachedUidGenerator extends DefaultUidGenerator implements DisposableBean {
private static final Logger LOGGER = LoggerFactory.getLogger(CachedUidGenerator.class);
private static final int DEFAULT_BOOST_POWER = 3;
/** Spring properties */
private int boostPower = DEFAULT_BOOST_POWER;
private int paddingFactor = RingBuffer.DEFAULT_PADDING_PERCENT;
private Long scheduleInterval;
private RejectedPutBufferHandler rejectedPutBufferHandler;
private RejectedTakeBufferHandler rejectedTakeBufferHandler;
/** RingBuffer */
private RingBuffer ringBuffer;
private BufferPaddingExecutor bufferPaddingExecutor;
@Override
public void afterPropertiesSet() throws Exception {
// initialize workerId & bitsAllocator
super.afterPropertiesSet();
// initialize RingBuffer & RingBufferPaddingExecutor
this.initRingBuffer();
LOGGER.info("Initialized RingBuffer successfully.");
}
@Override
public long getUID() {
try {
return ringBuffer.take();
} catch (Exception e) {
LOGGER.error("Generate unique id exception. ", e);
throw new UidGenerateException(e);
}
}
@Override
public String parseUID(long uid) {
return super.parseUID(uid);
}
@Override
public void destroy() throws Exception {
bufferPaddingExecutor.shutdown();
}
/**
* Get the UIDs in the same specified second under the max sequence
*
* @param currentSecond
* @return UID list, size of {@link BitsAllocator#getMaxSequence()} + 1
*/
protected List<Long> nextIdsForOneSecond(long currentSecond) {
// Initialize result list size of (max sequence + 1)
int listSize = (int) bitsAllocator.getMaxSequence() + 1;
List<Long> uidList = new ArrayList<>(listSize);
// Allocate the first sequence of the second, the others can be calculated with the offset
long firstSeqUid = bitsAllocator.allocate(currentSecond - epochSeconds, workerId, 0L);
for (int offset = 0; offset < listSize; offset++) {
uidList.add(firstSeqUid + offset);
}
return uidList;
}
/**
* Initialize RingBuffer & RingBufferPaddingExecutor
*/
private void initRingBuffer() {
// initialize RingBuffer
int bufferSize = ((int) bitsAllocator.getMaxSequence() + 1) << boostPower;
this.ringBuffer = new RingBuffer(bufferSize, paddingFactor);
LOGGER.info("Initialized ring buffer size:{}, paddingFactor:{}", bufferSize, paddingFactor);
// initialize RingBufferPaddingExecutor
boolean usingSchedule = (scheduleInterval != null);
this.bufferPaddingExecutor = new BufferPaddingExecutor(ringBuffer, this::nextIdsForOneSecond, usingSchedule);
if (usingSchedule) {
bufferPaddingExecutor.setScheduleInterval(scheduleInterval);
}
LOGGER.info("Initialized BufferPaddingExecutor. Using schdule:{}, interval:{}", usingSchedule, scheduleInterval);
// set rejected put/take handle policy
this.ringBuffer.setBufferPaddingExecutor(bufferPaddingExecutor);
if (rejectedPutBufferHandler != null) {
this.ringBuffer.setRejectedPutHandler(rejectedPutBufferHandler);
}
if (rejectedTakeBufferHandler != null) {
this.ringBuffer.setRejectedTakeHandler(rejectedTakeBufferHandler);
}
// fill in all slots of the RingBuffer
bufferPaddingExecutor.paddingBuffer();
// start buffer padding threads
bufferPaddingExecutor.start();
}
/**
* Setters for spring property
*/
public void setBoostPower(int boostPower) {
Assert.isTrue(boostPower > 0, "Boost power must be positive!");
this.boostPower = boostPower;
}
public void setRejectedPutBufferHandler(RejectedPutBufferHandler rejectedPutBufferHandler) {
Assert.notNull(rejectedPutBufferHandler, "RejectedPutBufferHandler can't be null!");
this.rejectedPutBufferHandler = rejectedPutBufferHandler;
}
public void setRejectedTakeBufferHandler(RejectedTakeBufferHandler rejectedTakeBufferHandler) {
Assert.notNull(rejectedTakeBufferHandler, "RejectedTakeBufferHandler can't be null!");
this.rejectedTakeBufferHandler = rejectedTakeBufferHandler;
}
public void setScheduleInterval(long scheduleInterval) {
Assert.isTrue(scheduleInterval > 0, "Schedule interval must positive!");
this.scheduleInterval = scheduleInterval;
}
}
獲取uid
@Override
public long getUID() {
try {
return ringBuffer.take();
} catch (Exception e) {
LOGGER.error("Generate unique id exception. ", e);
throw new UidGenerateException(e);
}
}
RingBuffer緩存已生成的id
(注意:這裏的RingBuffer不是Disruptor框架中的RingBuffer,但是藉助了很多Disruptor中RingBuffer的設計思想,比如使用緩存行填充解決僞共享問題)
RingBuffer爲環形數組,默認容量爲sequence可容納的最大值(8192個),可以通過boostPower參數設置大小。
tail指針、Cursor指針用於環形數組上讀寫slot:
-
Tail指針
表示Producer生產的最大序號(此序號從0開始,持續遞增)。Tail不能超過Cursor,即生產者不能覆蓋未消費的slot。當Tail已趕上curosr,此時可通過rejectedPutBufferHandler指定PutRejectPolicy -
Cursor指針
表示Consumer消費到的最小序號(序號序列與Producer序列相同)。Cursor不能超過Tail,即不能消費未生產的slot。當Cursor已趕上tail,此時可通過rejectedTakeBufferHandler指定TakeRejectPolicy
CachedUidGenerator採用了雙RingBuffer,Uid-RingBuffer用於存儲Uid、Flag-RingBuffer用於存儲Uid狀態(是否可填充、是否可消費)
由於數組元素在內存中是連續分配的,可最大程度利用CPU cache以提升性能。但同時會帶來「僞共享」FalseSharing問題,爲此在Tail、Cursor指針、Flag-RingBuffer中採用了CacheLine 補齊方式。
public class RingBuffer {
private static final Logger LOGGER = LoggerFactory.getLogger(RingBuffer.class);
/** Constants */
private static final int START_POINT = -1;
private static final long CAN_PUT_FLAG = 0L; //用於標記當前slot的狀態,表示可以put一個id進去
private static final long CAN_TAKE_FLAG = 1L; //用於標記當前slot的狀態,表示可以take一個id
public static final int DEFAULT_PADDING_PERCENT = 50; //用於控制何時填充slots的默認閾值:當剩餘的可用的slot的個數,小於bufferSize的50%時,需要生成id將slots填滿
/** The size of RingBuffer's slots, each slot hold a UID */
private final int bufferSize; //slots的大小,默認爲sequence可容量的最大值,即8192個
private final long indexMask;
private final long[] slots; //slots用於緩存已經生成的id
private final PaddedAtomicLong[] flags; //flags用於存儲id的狀態(是否可填充、是否可消費)
/** Tail: last position sequence to produce */
//Tail指針
//表示Producer生產的最大序號(此序號從0開始,持續遞增)。Tail不能超過Cursor,即生產者不能覆蓋未消費的slot。當Tail已趕上curosr,此時可通過rejectedPutBufferHandler指定PutRejectPolicy
private final AtomicLong tail = new PaddedAtomicLong(START_POINT); //
/** Cursor: current position sequence to consume */
//表示Consumer消費到的最小序號(序號序列與Producer序列相同)。Cursor不能超過Tail,即不能消費未生產的slot。當Cursor已趕上tail,此時可通過rejectedTakeBufferHandler指定TakeRejectPolicy
private final AtomicLong cursor = new PaddedAtomicLong(START_POINT);
/** Threshold for trigger padding buffer*/
private final int paddingThreshold; //用於控制何時填充slots的閾值
/** Reject put/take buffer handle policy */
//當slots滿了,無法繼續put時的處理策略。默認實現:無法進行put,僅記錄日誌
private RejectedPutBufferHandler rejectedPutHandler = this::discardPutBuffer;
//當slots空了,無法繼續take時的處理策略。默認實現:僅拋出異常
private RejectedTakeBufferHandler rejectedTakeHandler = this::exceptionRejectedTakeBuffer;
/** Executor of padding buffer */
//用於運行【生成id將slots填滿】任務
private BufferPaddingExecutor bufferPaddingExecutor;
RingBuffer填充時機
-
程序啓動時,將RingBuffer填充滿,緩存着8192個id
-
在調用getUID()獲取id時,檢測到RingBuffer中的剩餘id個數小於總個數的50%,將RingBuffer填充滿,使其緩存8192個id
-
定時填充(可配置是否使用以及定時任務的週期)
填充RingBuffer
/**
* Padding buffer fill the slots until to catch the cursor
*/
public void paddingBuffer() {
LOGGER.info("Ready to padding buffer lastSecond:{}. {}", lastSecond.get(), ringBuffer);
// is still running
if (!running.compareAndSet(false, true)) {
LOGGER.info("Padding buffer is still running. {}", ringBuffer);
return;
}
// fill the rest slots until to catch the cursor
boolean isFullRingBuffer = false;
while (!isFullRingBuffer) {
//獲取生成的id,放到RingBuffer中。
List<Long> uidList = uidProvider.provide(lastSecond.incrementAndGet());
for (Long uid : uidList) {
isFullRingBuffer = !ringBuffer.put(uid);
if (isFullRingBuffer) {
break;
}
}
}
// not running now
running.compareAndSet(true, false);
LOGGER.info("End to padding buffer lastSecond:{}. {}", lastSecond.get(), ringBuffer);
}
生成id(上面代碼中的uidProvider.provide調用的就是這個方法)
/**
* Get the UIDs in the same specified second under the max sequence
*
* @param currentSecond
* @return UID list, size of {@link BitsAllocator#getMaxSequence()} + 1
*/
protected List<Long> nextIdsForOneSecond(long currentSecond) {
// Initialize result list size of (max sequence + 1)
int listSize = (int) bitsAllocator.getMaxSequence() + 1;
List<Long> uidList = new ArrayList<>(listSize);
// Allocate the first sequence of the second, the others can be calculated with the offset
//這裏的實現很取巧
//因爲1秒內生成的id是連續的,所以利用第1個id來生成後面的id,而不用頻繁調用snowflake算法
long firstSeqUid = bitsAllocator.allocate(currentSecond - epochSeconds, workerId, 0L);
for (int offset = 0; offset < listSize; offset++) {
uidList.add(firstSeqUid + offset);
}
return uidList;
}
填充緩存行解決“僞共享”
關於僞共享,可以參考這篇文章《僞共享(false sharing),併發編程無聲的性能殺手》
//數組在物理上是連續存儲的,flags數組用來保存id的狀態(是否可消費、是否可填充),在填入id和消費id時,會被頻繁的修改。
//如果不進行緩存行填充,會導致頻繁的緩存行失效,直接從內存中讀數據。
private final PaddedAtomicLong[] flags;
//tail和cursor都使用緩存行填充,是爲了避免tail和cursor落到同一個緩存行上。
/** Tail: last position sequence to produce */
private final AtomicLong tail = new PaddedAtomicLong(START_POINT);
/** Cursor: current position sequence to consume */
private final AtomicLong cursor = new PaddedAtomicLong(START_POINT)
/**
* Represents a padded {@link AtomicLong} to prevent the FalseSharing problem<p>
*
* The CPU cache line commonly be 64 bytes, here is a sample of cache line after padding:<br>
* 64 bytes = 8 bytes (object reference) + 6 * 8 bytes (padded long) + 8 bytes (a long value)
* @author yutianbao
*/
public class PaddedAtomicLong extends AtomicLong {
private static final long serialVersionUID = -3415778863941386253L;
/** Padded 6 long (48 bytes) */
public volatile long p1, p2, p3, p4, p5, p6 = 7L;
/**
* Constructors from {@link AtomicLong}
*/
public PaddedAtomicLong() {
super();
}
public PaddedAtomicLong(long initialValue) {
super(initialValue);
}
/**
* To prevent GC optimizations for cleaning unused padded references
*/
public long sumPaddingToPreventOptimization() {
return p1 + p2 + p3 + p4 + p5 + p6;
}
}