RocketMQ存储机制01-存储文件组织与内存映射
Posted xuxiaojian
tags:
篇首语:本文由小常识网(cha138.com)小编为大家整理,主要介绍了RocketMQ存储机制01-存储文件组织与内存映射相关的知识,希望对你有一定的参考价值。
目录
上图是以CommitLog文件为例,展示了commitlog文件与MappedFile、MapppedFileQueue的关系。
你可以把磁盘里面commitlog文件夹下每个文件对应成MappedFile,而这个文件夹对应成MappedFileQueue。
先从MappedFileQueue看起
MappedFileQueue
private final String storePath;//存储目录
private final int mappedFileSize;//一个存储文件的大小
private final CopyOnWriteArrayList<MappedFile> mappedFiles = new CopyOnWriteArrayList<MappedFile>();//写时复制MappedFile列表,按顺序存储各个commitlog对于的MappedFile
private final AllocateMappedFileService allocateMappedFileService;//分配MappedFile服务
private long flushedWhere = 0;//当前刷盘指针的位置,表示之前的数据已经刷到磁盘
private long committedWhere = 0;//当前数据提交指针
MappedFile findMappedFileByOffset(final long offset, final boolean returnFirstOnNotFound)
/*
此处得到的第一个MappedFile的文件起始offset(即文件名)不一定是0,
之前的文件有可能已经被清除了。
*/
MappedFile mappedFile = this.getFirstMappedFile();
if (mappedFile != null) {
/*
之前的文件有可能已经被清除了(从this.mappedFiles里也会删掉)。因此不能直接用offset / this.mappedFileSize
计算offset对应的文件索引。
*/
int index = (int) ((offset / this.mappedFileSize) - (mappedFile.getFileFromOffset() / this.mappedFileSize));
if (index < 0 || index >= this.mappedFiles.size()) {
LOG_ERROR.warn("Offset for {} not matched. Request offset: {}, index: {}, " +
"mappedFileSize: {}, mappedFiles count: {}",
mappedFile,
offset,
index,
this.mappedFileSize,
this.mappedFiles.size());
}
try {
return this.mappedFiles.get(index);
} catch (Exception e) {
if (returnFirstOnNotFound) {
return mappedFile;
}
LOG_ERROR.warn("findMappedFileByOffset failure. ", e);
}
}
MappedFile
public static final int OS_PAGE_SIZE = 1024 * 4;//系统页缓存大小
protected static final Logger log = LoggerFactory.getLogger(LoggerName.STORE_LOGGER_NAME);
private static final AtomicLong TOTAL_MAPPED_VIRTUAL_MEMORY = new AtomicLong(0);//当前JVM实例中MappedFile虚拟内存
private static final AtomicInteger TOTAL_MAPPED_FILES = new AtomicInteger(0);//当前JVM实例中MappedFile对象个数
protected final AtomicInteger wrotePosition = new AtomicInteger(0);//当前文件的写指针
//ADD BY ChenYang
protected final AtomicInteger committedPosition = new AtomicInteger(0);//当前文件的提交指针
private final AtomicInteger flushedPosition = new AtomicInteger(0);//刷写磁盘的指针
protected int fileSize;//文件大小
protected FileChannel fileChannel;//文件通道
/**
* Message will put to here first, and then reput to FileChannel if writeBuffer is not null.
*/
protected ByteBuffer writeBuffer = null;//该buffer从transientStorePool申请,消息首先会放到这里,然后再提交到FileChannel
protected TransientStorePool transientStorePool = null;//堆外内存池,与上面的writeBuffer共同起作用
private String fileName;//文件名称
private long fileFromOffset;//文件起始物理偏移地址
private File file;//文件本身
private MappedByteBuffer mappedByteBuffer;//FileChannel对应的内存映射
private volatile long storeTimestamp = 0;//文件最后一次内容写入时间
private boolean firstCreateInQueue = false;//是否是MappedFileQueue中的第一个文件
init(final String fileName, final int fileSize)
this.fileName = fileName;
this.fileSize = fileSize;
this.file = new File(fileName);
this.fileFromOffset = Long.parseLong(this.file.getName());//文件名即为文件的起始物理偏移量
boolean ok = false;
ensureDirOK(this.file.getParent());//确保文件的父目录存在
try {
this.fileChannel = new RandomAccessFile(this.file, "rw").getChannel();//创建文件读写通道
this.mappedByteBuffer = this.fileChannel.map(MapMode.READ_WRITE, 0, fileSize);//获得映射buffer
TOTAL_MAPPED_VIRTUAL_MEMORY.addAndGet(fileSize);//增加
TOTAL_MAPPED_FILES.incrementAndGet();//增加
ok = true;
}
int commit(final int commitLeastPages)
if (writeBuffer == null) {
//no need to commit data to file channel, so just regard wrotePosition as committedPosition.
//没开启堆外内存池,不需要提交数据到fileChannel
return this.wrotePosition.get();//返回当前写的位置
}
if (this.isAbleToCommit(commitLeastPages)) {
if (this.hold()) {//???
commit0(commitLeastPages);
this.release();//???
} else {
log.warn("in commit, hold failed, commit offset = " + this.committedPosition.get());
}
}
// All dirty data has been committed to FileChannel.
//所有数据已经被提交到了FileChannel
if (writeBuffer != null && this.transientStorePool != null && this.fileSize == this.committedPosition.get()) {
this.transientStorePool.returnBuffer(writeBuffer);//将writeBuffer归还给transientStorePool
this.writeBuffer = null;
}
return this.committedPosition.get();//返回当前提交的位置
boolean isAbleToCommit(final int commitLeastPages)
int flush = this.committedPosition.get();
int write = this.wrotePosition.get();
if (this.isFull()) {
return true;//当前文件已经写满;,可以提交
}
if (commitLeastPages > 0) {
//本次要提交的数据页大于等于允许提交的最小阈值,可以提交
return ((write / OS_PAGE_SIZE) - (flush / OS_PAGE_SIZE)) >= commitLeastPages;
}
return write > flush;//当前写的位置大于提交的位置,可以提交
void commit0(final int commitLeastPages)
int writePos = this.wrotePosition.get();
int lastCommittedPosition = this.committedPosition.get();
if (writePos - this.committedPosition.get() > 0) {
try {
ByteBuffer byteBuffer = writeBuffer.slice();//创建一个buffer,与writeBuffer指向同一缓存区
byteBuffer.position(lastCommittedPosition);//回退buffer当前指针位置为lastCommittedPosition
byteBuffer.limit(writePos);//设置当前最大有效数据指针
this.fileChannel.position(lastCommittedPosition);
this.fileChannel.write(byteBuffer);
this.committedPosition.set(writePos);
} catch (Throwable e) {
log.error("Error occurred when commit data to FileChannel.", e);
}
}
以上是关于RocketMQ存储机制01-存储文件组织与内存映射的主要内容,如果未能解决你的问题,请参考以下文章