—> etcdmain/etcd.go startEtcdOrProxyV2():
—> etcdmain/config.go cfg := newConfig(): etcd初始化对象配置
—> etcdmain/config.go cfg.parse(Args[1:]): 解析etcd启动参数
—> etcdmain/etcd.go setupLogging(cfg)
—> etcdmain/etcd.go which := identifyDataDirOrDie(cfg.Dir) 分析Dir参数，解析etcd启动方式，如果Dir参数为空，也会走下面2个分支
—> case dirMember:
—> etcdmain/etcd.go startEtcd(&cfg.Config)
—> embed/etcd.go e, err := embed.StartEtcd(cfg) 启动etcd服务
—> embed/config.go cfg.Validate() 验证config值
—> embed/etcd.go e.Peers := startPeerListeners(cfg)
—> plns = make([]net.Listener, len(cfg.LPUrls))
—> for i, u := range cfg.LPUrls
—> pkg/transport/listener.go tlscfg = cfg.PeerTLSInfo.ServerConfig() CA file config
—> rafthttp/util.go plns[i] = rafthttp.NewListener(u, tlscfg)
—> transport/timeout_listener.go transport.NewTimeoutListener(u.Host, u.Scheme, tlscfg, ConnReadTimeout, ConnWriteTimeout)
—> pkg/transport/listener.go ln := newListener(u.Host, u.Scheme)
—> unix_listener.go if (u.Scheme == “unix” | “unixs”) return NewUnixListener(u.Host) unix socket via unix://laddr
—> dial.go else return net.Listen(“tcp”, u.Host) tcp socket
—> transport/timeout_listener.go ln = &rwTimeoutListener{}
—> pkg/transport/listener.go ln = wrapTLS(u.Host, u.Scheme, tlscfg, ln)
—> tls.go tlscfg.NewListener(ln, tlscfg) 创建listener，该listener接受inner连接 tls：Transport Layer Security 传输层安全协议
—> embed/etcd.go e.sctxs := startClientListeners(cfg)
—> if cfg.ClientAutoTLS && cfg.ClientTLSInfo.Empty() cfg.ClientTLSInfo = transport.SelfCert(path.Join(cfg.Dir, “fixtures/client"), cfg.LCurls) 配置tls信息
—> sctxs = make(map[string] *serveCtx)
—> sctx.l = net.Listen(proto, cfg.LCurls) proto=“tcp|unix|unixs" 创建socket 进行listen
—> sctx.l = transport.LimitListener(sctx.l, int(fdLimit - reservedInternalFDNum)) 设置listen limit上限值
—> pkg/transport/keepalive_listener.go sctx.l = transport.NewKeepAliveListener(sctx.l, “tcp”, cfg) 创建socket
—> newTLSKeepaliveListener(sctx.l, cfg)
—> e.Clients = append(e.Clients, e.sctxs.l)
—> srvcfg := &etcdserver.ServerConfig{cfg}
—> etcdserver/server.go e.Server = etcdserver.NewServer(srvcfg)
—> store/store.go st := store.New(StoreClusterPrefix, StoreKeysPrefix)
—> store/store.go s := newStore(namespaces ...)
—> s := new(store)
—> s.Root = newDir(s, “/“, s.CurrentIndex, Permanent)
—> for namespace := range namespaces s.Root.Add(newDir(s, namespace, s.CurrentIndex, s.Root, Permanent))
—> s.Stats = newStats()
—> s.WatcherHub = newWatchHub(1000)
—> s.ttlKeyHeap = newTtlKeyHeap()
—> s.readonlySet = types.NewUnsafeSet(append(namespaces, “/“) ...)
—> s.clock = clockwork.NewRealClock()
—> haveWAL := wal.Exist(cfg.WALDir())
—> ss := snap.New(cfg.SnapDir())
—> bepath := path.Join(cfg.SnapDir(), databaseFilename)
—> mvcc/backend/backend.go be := backend.NewDefaultBackend(bepath)
—> newBackend(bepath, defaultBatchInterval, defaultBatchLimit)
—> db := bolt.Open(bepath, 0600, boltOpenOptions)
—> b := &backend{}
—> mvcc/backend/batch_tx.go b.batchTx = newBatchTx(b)
—> tx := &batchTx{b}
—> tx.Commit()
—> go b.run()
—> t := time.NewTimer(b.batchInterval)
—> for
—> select
—> case <- t,C:
—> case <- b.stopc:
—> b.batchTx.CommitAndStop()
—> b.batchTx.Commit()
—> b.Reset(b.batchInterval)
—> rafthttp/util.go prt := rafthttp.NewRoundTripper(cfg.PeerTLSInfo, cfg.peerDialTimeout())
—> pkg/transport/timeout_transport.go transport.NewTimoutTransport(cfg.PeerTLSInfo, cfg.peerDialTimeout())
—> tr := NewTransport(info, dialtimeoutd)
—> tr.Dial =(&rwTimeoutDialer{}).Dial
—> swith
—> case !haveWAL && !cfg.NewCluster
—> cfg.VerifyJoinExisting()
—> cl = membership.NewClusterFromURLsMap(cfg.InitialClusterToken, cfg.InitialPeerURLsMap)
—> existingCluster = GetClusterFromRemotePeers(getRemotePeerURLs(cl, cfg.Name), prt)
—> membership.ValidateClusterAndAssignIDs(cl, existingCluster)
—> isCompatibleWithCluster(cl, cl.MemberByName(cfg.Name).ID, prt)
—> remotes = existingCluster.Members()
—> cl.SetID(existingCluster.ID())
—> cl.SetStore(st)
—> cl.SetBackend(be)
—> cfg.Print()
—> id, n, s, w = startNode(cfg, cl)
—> case !haveWAL && cfg.NewCluster
—> cfg.VerifyBootstrape()
—> cl = membership.NewClusterFromURLsMap(cfg.InitialClusterToken, cfg.InitialPeerURLsMap)
—> m := cl.MemberByName(cfg.Name)
—> isMemberBootstrapped(cl, cfg.Name, prt, cfg.bootstrapTimeout())
—> if cfg.ShouldDiscover()
—> str = discovery.JoinCluster(cfg.DiscoveryURL, cfg.DiscoveryProxy, m.ID, cfg.InitialPeerURLsMap.String())
—> urlsmap = types.NewURLsMap(str)
—> checkDuplicateURL(urlsmap)
—> cl = membership.NewClusterFromURLsMap(cfg.InitoalClusterToken, urlsmap)
—> cl.SetStore(st)
—> cl.SetBackend(be)
—> cfg.PrintWithInitial()
—> id, n, s, w = startNode(cfg, cl, cl.MemberIDs())
—> case haveWAL:
—> fileutil.IsDirWriteable(cfg.MemberDir())
—> fileutil.IsDirWriteable(cfg.WALDir())
—> snapshot = ss.Load()
—> st.Recovery(snapshot.Data)
—> cfg.Print()
—> if !cfg.ForceNewCluster
—> id, cl, n, s, w = restartNode(cfg, snapshot)
—> else
—> id, cl, n, s, w = restartAsStandaloneNode(cfg, snapshot)
—> cl.SetStore(st)
—> cl.SetBackend(be)
—> cl.Recover()
—> sstats := &stats.ServerStats{}
—> sstats.Initialize()
—> lstats := stats.NewLeaderStats(id.String())
—> srv = &EtcdServer{}
—> srv.applyV2 = &applierV2store{}
—> srv.be = be
—> srv.lessor = lease.NewLessor(srv.be)
—> srv.kv = mvcc.New(src.be, srv.lessor, &srv.consistIndex)
—> srv.consistIndex.setConsistentIndex(srv.kv.ConsistentIndex())
—> srv.authStore = auth.NewAuthStore(srv.be)
—> h := cfg.AutoCompactionRetention
—> srv.compactor = compactor.NewPeriodic(h, srv,kv, srv)
—> tr := &rafthttp.Transport{}
—> tr.Start()
—> for m := range remotes tr.AddRemote(m.ID, m.PeerURLs)
—> for m := range cl.Members() tr.AddPeer(m.ID, m.PeerURLs)
—> srv.r.transport = tr
—> etcdserver/server.go e.Server.Start()
—> interrupt_unix.go osutil.RegisterInterruptHandler(e.Server.Stop) 向系统注册中断事件
—> case dirProxy:
—> etcdmain/etcd.go startProxy(&cfg.Config)
