unblocked. */
while (listLength(server.unblocked_clients)) {
ln = listFirst(server.unblocked_clients);
redisAssert(ln != NULL);
c = ln->value;
listDelNode(server.unblocked_clients,ln);
c->flags &= ~REDIS_UNBLOCKED;
/* Process remaining data in the input buffer. */
//处理客户端在阻塞期间接收到的客户端发送的请求
if (c->querybuf && sdslen(c->querybuf) > 0) {
server.current_client = c;
processInputBuffer(c);
server.current_client = NULL;
}
}
/* Write the AOF buffer on disk */
//将server.aof_buf中的数据追加到AOF文件中并fsync到硬盘上
flushAppendOnlyFile(0);
}
通过上面的代码及注释可以发现,beforeSleep函数做了三件事:1、处理过期键,2、处理阻塞期间的客户端请求,3、将server.aof_buf中的数据追加到AOF文件中并fsync刷新到硬盘上,flushAppendOnlyFile函数给定了一个参数force,表示是否强制写入AOF文件,0表示非强制即支持延迟写,1表示强制写入。
void flushAppendOnlyFile(int force) {
ssize_t nwritten;
int sync_in_progress = 0;
if (sdslen(server.aof_buf) == 0) return;
// 返回后台正在等待执行的 fsync 数量
if (server.aof_fsync == AOF_FSYNC_EVERYSEC)
sync_in_progress = bioPendingJobsOfType(REDIS_BIO_AOF_FSYNC) != 0;
// AOF 模式为每秒 fsync ,并且 force 不为 1 如果可以的话,推延冲洗
if (server.aof_fsync == AOF_FSYNC_EVERYSEC && !force) {
/* With this append fsync policy we do background fsyncing.
* If the fsync is still in progress we can try to delay
* the write for a couple of seconds. */
// 如果 aof_fsync 队列里已经有正在等待的任务
if (sync_in_progress) {
// 上一次没有推迟冲洗过,记录推延的当前时间,然后返回
if (server.aof_flush_postponed_start == 0) {
/* No previous write postponinig, remember that we are
* postponing the flush and return. */
server.aof_flush_postponed_start = server.unixtime;
return;
} else if (server.unixtime - server.aof_flush_postponed_start < 2) {
// 允许在两秒之内的推延冲洗
/* We were already waiting for fsync to finish, but for less
* than two seconds this is still ok. Postpone again. */
return;
}
/* Otherwise fall trough, and go write since we can't wait
* over two seconds. */
server.aof_delayed_fsync++;
redisLog(REDIS_NOTICE,"Asynchronous AOF fsync is taking too long (disk is busy ). Writing the AOF buffer without waiting for fsync to complete, this may slow down Redis.");
}
}
/* If you are following this code path, then we are going to write so
* set reset the postponed flush sentinel to zero. */
server.aof_flush_postponed_start = 0;
/* We want to perform a single write. This should be guaranteed atomic
* at least if the filesystem we are writing is a real physical one.
* While this will save us against the server being killed I don't think
* there is much to do about the whole server stopping for power problems
* or alike */
// 将 AOF 缓存写入到文件,如果一切幸运的话,写入会原子性地完成
nwritten = write(server.aof_fd,server.aof_buf,sdslen(server.aof_buf));
if (nwritten != (signed)sdslen(server.aof_buf)) {//出错
/* Ooops, we are in troubles. The best thing to do for now is
* aborting instead of giving the illusion that everything is
* working as expected. */
if (nwritten == -1) {
redisLog(REDIS_WARNING,"Exiting on error writing to the append-only file: %s",strerror(errno));
} else {
redisLog(REDIS_WARNING,"Exiting on short write while writing to "
"the append-only file: %s (nwritten=%ld, "
"expected=%ld)",
strerror(errno),
(long)nwritten,
(long)sdslen(server.aof_buf));
if (ftruncate(server.aof_fd, server.aof_current_size) == -1) {
redisLog(REDIS_WARNING, "Could not remove short write "
"from the append-only file. Redis may refuse "
"to load the AOF the next time it starts. "
"ftruncate: %s", strerror(errno));
}
}
exit(1);
}
server.aof_current_size += nwritten;
/* Re-use AOF buffer when it is small enough. The maximum comes from the
* arena size of 4k minus some overhead (but is otherwise arbitrary). */
// 如果 aof 缓存不是太大,那么重用它,否则,清空 aof 缓存
if ((sdslen(server.aof_buf)+sdsavail(server.aof_buf)) < 4000) {
sdsclear(server.aof_buf);
} else {
sdsfree(server.aof_buf);
server.aof_buf = sdsempty();
}
/* Don't fsync if no-appendfsync-on-rewrite is set to yes and there are
* children doing I/O in the background. */
//aof rdb子进程运行中不支持fsync并且aof rdb子进程正在运行,那么直接返回,
//但是数据已经写到aof文件中,只是没有刷新到硬盘
if (server.aof_no_fsync_on_rewrite &&
(server.aof_child_pid != -1 || server.rdb_child_pid != -1))
return;
/* Perform the fsync if needed. */
if (server.aof_fsync == AOF_FSYNC_ALWAYS) {//总是fsync,那么直接进行fsync
/* aof_fsync is defined as fdatasync() for Linux in order to avoid
* flushing metadata. */
aof_fsync(server.aof_fd); /* Let's try to get this data on the disk */
server.aof_last_fsync = server.unixtime;
} else if ((server.aof_fsync == AOF_FSYNC_EVERYSEC &&
server.unixtime > server.aof_last_fsync)) {
if (!sync_in_progress) aof_background_fsync(server.aof_fd);//放到后台线程进行fsync
server.aof_last_fsync = server.unixtime;
}
}
上述代码中请关注server.aof_fsync参数,即设置Redis fsync AOF文件到硬盘的策略,如果设置为AOF_FSYNC_ALWAYS,那么直接在主进程中fsync,如果设置为AOF_FSYNC_EVERYSEC,那么放入后台线程中fsync,后台线程的代码在bio.c中。
小结
文章写到这,已经解决的了Redis Server启动加载AOF文件和如何将客户端请求产生的新的数据追加到AOF文件中,对于追加数据到AOF文件中,根据fsync的配置策略如何将写入到AOF文件中的新数据刷新到硬盘中,直接在主进程中fsync或是在后台线程fsync。
至此,AOF数据持久化还剩下如何rewrite AOF,接受客户端发送的BGREWRITEAOF请求,此部分内容待下篇博客中解析。
感谢此篇博客给我在理解Redis AOF数据持久化方面的巨大帮助,http://chenzhenianqing.cn/articles/786.html
本人Redis-2.8.2的源码注释已经放到Github中,有需要的读者可以下载,我也会在后续的时间中更新,https://github.com/xkeyideal/annotated-redis-2.8.2
本人不怎么会使用Git,望有人能教我一下。