redis事务的实现

redis虽然是一个非关系型内存数据库， 但是它支持传统的关系型数据库的事务。redis事务提供了一种将多个命令请求打包，然后一次性、按照顺序地执行多个命令的机制，并且在事务执行的期间，服务器不会中断事务而去执行其他不在事务中的命令请求，它会把事务中所有的命令都执行完毕才会去执行其他的命令。

事务实现命令

Redis中提供了multi、discard、exec、watch、unwatch这几个命令来实现事务的功能。

序号	命令	说明
1	watch	监视一个(或多个) key ，如果在事务执行之前这个(或这些) key 被其他命令所改动，那么事务将被打断。
2	unwatch	取消 WATCH 命令对所有 key 的监视。
3	multi	标记一个事务块的开始。
4	exec	执行所有事务块内的命令。
5	discard	取消事务，放弃执行事务块内的所有命令。

redis事务相关的代码都放在multi.c里面， 接下来我们继续看一下这些命令是如何实现的。

watch、unwatch

准确的说， watch和unwatch不算是事务的命令， 但是它们可以检查某个事务是不是有被两个或者多个客户端修改过。如果没有被修改过，该事务就可以被执行， 否则就不能被执行。
用法：

watch key1 key2 key3

void watchCommand(client *c) {
    int j;

    if (c->flags & CLIENT_MULTI) {
        addReplyError(c,"WATCH inside MULTI is not allowed");
        return;
    }
    // 对每一个要监控的对象， 调用watchForKey。
    for (j = 1; j < c->argc; j++)
        watchForKey(c,c->argv[j]);
    addReply(c,shared.ok);
}

void unwatchCommand(client *c) {
    unwatchAllKeys(c);
    c->flags &= (~CLIENT_DIRTY_CAS);
    addReply(c,shared.ok);
}

void watchForKey(client *c, robj *key) {
    list *clients = NULL;
    listIter li;
    listNode *ln;
    watchedKey *wk;

    /* 检查要监控的Key是不是已经存在 */
    listRewind(c->watched_keys,&li);
    while((ln = listNext(&li))) {
        wk = listNodeValue(ln);
        if (wk->db == c->db && equalStringObjects(key,wk->key))
            return; /* Key already watched */
    }
    /* 把key添加到watched_keys */
    clients = dictFetchValue(c->db->watched_keys,key);
    if (!clients) {
        clients = listCreate();
        dictAdd(c->db->watched_keys,key,clients);
        incrRefCount(key);
    }
    listAddNodeTail(clients,c);
    /* Add the new key to the list of keys watched by this client */
    wk = zmalloc(sizeof(*wk));
    wk->key = key;
    wk->db = c->db;
    incrRefCount(key);
    listAddNodeTail(c->watched_keys,wk);
}

void unwatchAllKeys(client *c) {
    listIter li;
    listNode *ln;

    if (listLength(c->watched_keys) == 0) return;
    listRewind(c->watched_keys,&li);
    while((ln = listNext(&li))) {
        list *clients;
        watchedKey *wk;

        /* Lookup the watched key -> clients list and remove the client
         * from the list */
        wk = listNodeValue(ln);
        clients = dictFetchValue(wk->db->watched_keys, wk->key);
        serverAssertWithInfo(c,NULL,clients != NULL);
        listDelNode(clients,listSearchKey(clients,c));
        /* Kill the entry at all if this was the only client */
        if (listLength(clients) == 0)
            dictDelete(wk->db->watched_keys, wk->key);
        /* Remove this watched key from the client->watched list */
        listDelNode(c->watched_keys,ln);
        decrRefCount(wk->key);
        zfree(wk);
    }
}

multi

multi 命令是一个事务开始的标志， 表示从该命令之后， exec开始之前， 除了exec， discard，multi或者watch命令之外， 所有的其他command都会被放入一个command的队列里面， 该队列是一个multiCmd类型的队列， 对于每一个command都会生成一个multiCmd对象。
该命令本身的执行比较简单， 就是将一个redisClient的CLIENT_MULTI flag设定。以便其后的command可以知道需要加入队列， 而不是直接执行。

void multiCommand(client *c) {
    if (c->flags & CLIENT_MULTI) {
        addReplyError(c,"MULTI calls can not be nested");
        return;
    }
    c->flags |= CLIENT_MULTI;
    addReply(c,shared.ok);
}
int processCommand(client *c) {
     if (c->flags & CLIENT_MULTI &&
        c->cmd->proc != execCommand && c->cmd->proc != discardCommand &&
        c->cmd->proc != multiCommand && c->cmd->proc != watchCommand)
    {
        queueMultiCommand(c);
        addReply(c,shared.queued);
    } 
}
void queueMultiCommand(client *c) {
    multiCmd *mc;
    int j;

    c->mstate.commands = zrealloc(c->mstate.commands,
            sizeof(multiCmd)*(c->mstate.count+1));
    mc = c->mstate.commands+c->mstate.count;
    mc->cmd = c->cmd;
    mc->argc = c->argc;
    mc->argv = zmalloc(sizeof(robj*)*c->argc);
    memcpy(mc->argv,c->argv,sizeof(robj*)*c->argc);
    for (j = 0; j < c->argc; j++)
        incrRefCount(mc->argv[j]);
    c->mstate.count++;
}

exec

该命令与multi是成对出现的，类似于一个事务的commit， 在执行事务的command的时候， 是将multiCmd队列里面的每一条command逐个执行。

void execCommand(client *c) {
    int j;
    robj **orig_argv;
    int orig_argc;
    struct redisCommand *orig_cmd;
    int must_propagate = 0; /* Need to propagate MULTI/EXEC to AOF / slaves? */
    int was_master = server.masterhost == NULL;

    if (!(c->flags & CLIENT_MULTI)) {
        addReplyError(c,"EXEC without MULTI");
        return;
    }

    /* Check if we need to abort the EXEC because:
     * 1) Some WATCHed key was touched.
     * 2) There was a previous error while queueing commands.
     * A failed EXEC in the first case returns a multi bulk nil object
     * (technically it is not an error but a special behavior), while
     * in the second an EXECABORT error is returned. */
    if (c->flags & (CLIENT_DIRTY_CAS|CLIENT_DIRTY_EXEC)) {
        addReply(c, c->flags & CLIENT_DIRTY_EXEC ? shared.execaborterr :
                                                  shared.nullmultibulk);
        discardTransaction(c);
        goto handle_monitor;
    }

    /* Exec all the queued commands */
    unwatchAllKeys(c); /* Unwatch ASAP otherwise we'll waste CPU cycles */
    orig_argv = c->argv;
    orig_argc = c->argc;
    orig_cmd = c->cmd;
    addReplyMultiBulkLen(c,c->mstate.count);
    for (j = 0; j < c->mstate.count; j++) {
        c->argc = c->mstate.commands[j].argc;
        c->argv = c->mstate.commands[j].argv;
        c->cmd = c->mstate.commands[j].cmd;

        /* Propagate a MULTI request once we encounter the first command which
         * is not readonly nor an administrative one.
         * This way we'll deliver the MULTI/..../EXEC block as a whole and
         * both the AOF and the replication link will have the same consistency
         * and atomicity guarantees. */
        if (!must_propagate && !(c->cmd->flags & (CMD_READONLY|CMD_ADMIN))) {
            execCommandPropagateMulti(c);
            must_propagate = 1;
        }

        call(c,CMD_CALL_FULL);

        /* Commands may alter argc/argv, restore mstate. */
        c->mstate.commands[j].argc = c->argc;
        c->mstate.commands[j].argv = c->argv;
        c->mstate.commands[j].cmd = c->cmd;
    }
    c->argv = orig_argv;
    c->argc = orig_argc;
    c->cmd = orig_cmd;
    discardTransaction(c);

    /* Make sure the EXEC command will be propagated as well if MULTI
     * was already propagated. */
    if (must_propagate) {
        int is_master = server.masterhost == NULL;
        server.dirty++;
        /* If inside the MULTI/EXEC block this instance was suddenly
         * switched from master to slave (using the SLAVEOF command), the
         * initial MULTI was propagated into the replication backlog, but the
         * rest was not. We need to make sure to at least terminate the
         * backlog with the final EXEC. */
        if (server.repl_backlog && was_master && !is_master) {
            char *execcmd = "*1\r\n$4\r\nEXEC\r\n";
            feedReplicationBacklog(execcmd,strlen(execcmd));
        }
    }

handle_monitor:
    /* Send EXEC to clients waiting data from MONITOR. We do it here
     * since the natural order of commands execution is actually:
     * MUTLI, EXEC, ... commands inside transaction ...
     * Instead EXEC is flagged as CMD_SKIP_MONITOR in the command
     * table, and we do it here with correct ordering. */
    if (listLength(server.monitors) && !server.loading)
        replicationFeedMonitors(c,server.monitors,c->db->id,c->argv,c->argc);
}

discard

discard是将所有queue的command的丢弃掉。

void discardCommand(client *c) {
    if (!(c->flags & CLIENT_MULTI)) {
        addReplyError(c,"DISCARD without MULTI");
        return;
    }
    discardTransaction(c);
    addReply(c,shared.ok);
}
void discardTransaction(client *c) {
    freeClientMultiState(c);
    initClientMultiState(c);
    c->flags &= ~(CLIENT_MULTI|CLIENT_DIRTY_CAS|CLIENT_DIRTY_EXEC);
    unwatchAllKeys(c);
}