本文基于社区版Redis 4.0.8
Redis服务器接收到的命令请求首先存储在客户端对象的querybuf输入缓冲区,然后解析命令请求的各个参数,并存储在客户端对象的argv和argc字段。
客户端解析命令请求的入口函数为readQueryFromClient,会读取socket数据存储到客户端对象的输入缓冲区,并调用函数processInputBuffer解析命令请求。
注:内联命令:使用telnet会话输入命令的方式
void processInputBuffer(client *c) { ...... //循环遍历输入缓冲区,获取命令参数,调用processMultibulkBuffer解析命令参数和长度 while(sdslen(c->querybuf)) { if (c->reqtype == PROTO_REQ_INLINE) { if (processInlineBuffer(c) != C_OK) break;//处理telnet方式的内联命令 } else if (c->reqtype == PROTO_REQ_MULTIBULK) { if (processMultibulkBuffer(c) != C_OK) break; //解析命令参数和长度暂存到客户端结构体中 } else { serverPanic("Unknown request type"); } } } //解析命令参数和长度暂存到客户端结构体中 int processMultibulkBuffer(client *c) { //定位到行尾 newline = strchr(c->querybuf,'\r'); //解析命令请求参数数目,并存储在客户端对象的c->multibulklen字段 serverAssertWithInfo(c,NULL,c->querybuf[0] == '*'); ok = string2ll(c->querybuf+1,newline-(c->querybuf+1),&ll); c->multibulklen = ll; pos = (newline-c->querybuf)+2;//记录已解析命令的请求长度resp的长度 /* Setup argv array on client structure */ //分配请求参数存储空间 c->argv = zmalloc(sizeof(robj*)*c->multibulklen); // 开始循环解析每个请求参数 while(c->multibulklen) { ...... newline = strchr(c->querybuf+pos,'\r'); if (c->querybuf[pos] != '$') { return C_ERR; ok = string2ll(c->querybuf+pos+1,newline-(c->querybuf+pos+1),&ll); pos += newline-(c->querybuf+pos)+2; c->bulklen = ll;//字符串参数长度暂存在客户端对象的bulklen字段 //读取该长度的参数内容,并创建字符串对象,同时更新待解析参数multibulklen c->argv[c->argc++] =createStringObject(c->querybuf+pos,c->bulklen); pos += c->bulklen+2; c->multibulklen--; }
当multibulklen的值更新为0时,表示参数解析完成,开始调用processCommand来处理命令,处理命令前有很多校验逻辑,如下:
void processInputBuffer(client *c) { ...... //调用processCommand来处理命令 if (processCommand(c) == C_OK) { ...... } } //处理命令函数 int processCommand(client *c) { //校验是否是quit命令 if (!strcasecmp(c->argv[0]->ptr,"quit")) { addReply(c,shared.ok); c->flags |= CLIENT_CLOSE_AFTER_REPLY; return C_ERR; } //调用lookupCommand,查看该命令是否存在 c->cmd = c->lastcmd = lookupCommand(c->argv[0]->ptr); if (!c->cmd) { flagTransaction(c); addReplyErrorFormat(c,"unknown command '%s'", (char*)c->argv[0]->ptr); return C_OK; //检查用户权限 if (server.requirepass && !c->authenticated && c->cmd->proc != authCommand) { addReply(c,shared.noautherr); //还有很多检查,不一一列举,比如集群/持久化/复制等 /* 真正执行命令 */ if (c->flags & CLIENT_MULTI && c->cmd->proc != execCommand && c->cmd->proc != discardCommand && c->cmd->proc != multiCommand && c->cmd->proc != watchCommand) queueMultiCommand(c); //将结果写入outbuffer addReply(c,shared.queued); } // 调用execCommand执行命令 void execCommand(client *c) { call(c,CMD_CALL_FULL);//调用call执行命令 //调用execCommand调用call执行命令 void call(client *c, int flags) { start = ustime(); c->cmd->proc(c);//执行命令 duration = ustime()-start; //如果是慢查询,记录慢查询 if (flags & CMD_CALL_SLOWLOG && c->cmd->proc != execCommand) { char *latency_event = (c->cmd->flags & CMD_FAST) ? "fast-command" : "command"; latencyAddSampleIfNeeded(latency_event,duration/1000); //记录到慢日志中 slowlogPushEntryIfNeeded(c,c->argv,c->argc,duration); //更新统计信息:当前命令执行时间和调用次数 if (flags & CMD_CALL_STATS) { c->lastcmd->microseconds += duration; c->lastcmd->calls++;
Redis返回结果并不是直接返回给客户端,而是先写入到输出缓冲区(buf字段)或者输出链表(reply字段)
int processCommand(client *c) { ...... //将结果写入outbuffer addReply(c,shared.queued); ...... } //将结果写入outbuffer void addReply(client *c, robj *obj) { //调用listAddNodeHead将客户端添加到服务端结构体的client_pending_write链表,以便后续能快速查找出哪些客户端有数据需要发送 if (prepareClientToWrite(c) != C_OK) return; //然后添加字符串到输出缓冲区 if (_addReplyToBuffer(c,obj->ptr,sdslen(obj->ptr)) != C_OK) //如果添加失败,则添加到输出链表中 _addReplyObjectToList(c,obj); }
addReply函数只是将待发送给客户端的数据暂存在输出链表或者输出缓冲区,那么什么时候将这些数据发送给客户端呢?答案是开启事件循环时,调用的beforesleep函数,该函数专门执行一些不是很费时的操作,如过期键删除,向客户端返回命令回复等
void beforeSleep(struct aeEventLoop *eventLoop) { ...... /* Handle writes with pending output buffers. */ handleClientsWithPendingWrites(); } //回复客户端命令函数 int handleClientsWithPendingWrites(void) { listIter li; listNode *ln; int processed = listLength(server.clients_pending_write); listRewind(server.clients_pending_write,&li); while((ln = listNext(&li))) { client *c = listNodeValue(ln); c->flags &= ~CLIENT_PENDING_WRITE; listDelNode(server.clients_pending_write,ln); /* 发送客户端数据 */ if (writeToClient(c->fd,c,0) == C_ERR) continue; /* If there is nothing left, do nothing. Otherwise install * the write handler. */ //如果数据量很大,一次性没有发送完成,则进行添加文件事件,监听当前客户端socket文件描述符的可写事件即可 if (clientHasPendingReplies(c) && aeCreateFileEvent(server.el, c->fd, AE_WRITABLE, sendReplyToClient, c) == AE_ERR) { freeClientAsync(c); } } return processed;
到这里,命令请求才算真正处理完成了。