Ipvs-mode proxier

1. 概述

关于ipvs-mode proxier基础知识可参看官方文档(英文版中文版),其官方文档主要介绍以下几方面内容:

  1. ipvs技术简介和对比iptables-mode所带来的好处;
  2. ipvs-mode proxier按用户配置不同所生成的用户层iptables规则示例(masquerade-all/cluster-cidr/Load Balancer/NodePort/externalIPs);
  3. 如何kube-proxy运行ipvs模式、运行必要条件、运行debug和排错操作。

本文分析将聚焦在代码层的实现解析 (如运行时必要条件检测的代码实现是怎么样的?ipvs为实现service层是如何实现的?iptables规则代码是怎样的?proxier完整的实现逻辑与方式是怎样?等等) 。

ipvs proxy模式主要依赖几个底层技术如 ipvs/ipset /iptables/netlink(用户空间与内核空间异步通信机制),有必要预先对其基础用途或技术细节进行扩展知识的熟悉,将有助于对整个ipvs-mode proxier的实现更深层次的理解。

Ipvs-mode proxier使用ipvs NAT模式实现,ipvs集群操作(如虚拟服务器、RealServer)是通过netlink内核通迅创建标准的协议格式通迅消息体进行交互实现。 Ipvs-mode proxier也同样使用了iptables固定模板规则结合ipset集来进行动态管理变化更新。

Ipvs-mode proxier整个代码机制逻辑与iptables-mode一致(参看iptable-mode代码逻辑示意图)。同样是通过同步apiserver事件及更新信息,生成相应的路由规则。但ipvs-mode服务规则不同于iptables-mode,不仅使用了ipset扩展的方式简化iptables规则条目和优化性能,而且还使用ipvs技术实现更丰富的集群负载策略管理。其规则生成操作须对ipset集、iptables规则、ipvs集群进行同步更新操作,关键逻辑代码在syncProxyRules()内。

2. Ipvs-mode proxer 对象创建与初始化

ProxyServer实例化时初始化了proxier模式,如果代理模式指定为Ipvs,则创建proxier对象,且指定其service与endpoints的事件处理器。

cmd/kube-proxy/app/server_others.go:59

func newProxyServer(...) (*ProxyServer, error) {
           //...
  else if proxyMode == proxyModeIPVS {          //当proxy模式指定为IPVS模式(命令参数或配置文件)
        klog.V(0).Info("Using ipvs Proxier.")
        proxierIPVS, err := ipvs.NewProxier(        //创建ipvs-mode proxier对象
               //...
        )
    // porxyServer的proxier对象与事件处理器的指定
        proxier = proxierIPVS
        serviceEventHandler = proxierIPVS
        endpointsEventHandler = proxierIPVS
           //...
}

ipvs-mode proxier对象实例化NewProxier(),对ipvs环境进行初始化

相关内核参数调整说明:

  • net/ipv4/conf/all/route_localnet 是否允许外部访问localhost;
  • net/bridge/bridge-nf-call-iptables 为二层的网桥在转发包时也会被iptables的FORWARD规则所过滤,这样就会出现L3层的iptables rules去过滤L2的帧的问题;
  • net/ipv4/vs/conntrack 开启NFCT(Netfilter connection tracking连接与状态跟踪);
  • net/ipv4/vs/conn_reuse_mode 网络连接复用模式的选择;
  • net/ipv4/vs/expire_nodest_conn 值为0,当LVS转发数据包,发现目的RS无效(删除)时,会丢弃该数据包,但不删除相应连接。值为1时,则马上释放相应连接;
  • net/ipv4/vs/expire_quiescent_template 值为0,当RS的weight值=0(如,健康检测失败,应用程序将RS weight置0)时,会话保持的新建连接 还会继续调度到该RS上;值为1,则马上将会话保持的连接模板置为无效,重新调度新的RS。如果有会话保持的业务,建议该值配置为1;
  • net/ipv4/ip_forward 是否打开ipv4的IP转发模式;
  • net/ipv4/conf/all/arp_ignore 定义对目标地址为本地IP的ARP询问不同的应答模式(0~8),模式1表示:只回答目标IP地址是来访网络接口本地地址的ARP查询请求;
  • net/ipv4/conf/all/arp_announce 对网络接口上,本地IP地址的发出的,ARP回应,作出相应级别的限制;值为2表示:对查询目标使用最适当的本地地址;

pkg/proxy/ipvs/proxier.go:280

func NewProxier(...) (*Proxier, error) {
    // sysctl配置项 "net/ipv4/conf/all/route_localnet" 值为1
    if val, _ := sysctl.GetSysctl(sysctlRouteLocalnet); val != 1 {
        if err := sysctl.SetSysctl(sysctlRouteLocalnet, 1); err != nil {
            return nil, fmt.Errorf("can't set sysctl %s: %v", sysctlRouteLocalnet, err)
        }
    }
  //...
  // sysctl配置项 "net/bridge/bridge-nf-call-iptables"  值为1
  sysctl.GetSysctl(sysctlBridgeCallIPTables)
  // sysctl配置项 "net/ipv4/vs/conntrack" 值为1
  sysctl.SetSysctl(sysctlVSConnTrack, 1)
  // sysctl配置项 "net/ipv4/vs/conn_reuse_mode" 值为0
  sysctl.SetSysctl(sysctlConnReuse, 0)
  // sysctl配置项 "net/ipv4/vs/expire_nodest_conn" 值为1
  sysctl.SetSysctl(sysctlExpireNoDestConn, 1)
  // sysctl配置项 "net/ipv4/vs/expire_quiescent_template" 值为1
  sysctl.SetSysctl(sysctlExpireQuiescentTemplate, 1)
  // sysctl配置项 "net/ipv4/ip_forward" 值为1 
  sysctl.SetSysctl(sysctlForward, 1)
  // sysctl配置项 "net/ipv4/conf/all/arp_ignore" 值为1
  sysctl.SetSysctl(sysctlArpIgnore, 1)
  // sysctl配置项 "net/ipv4/conf/all/arp_announce" 值为2
  sysctl.SetSysctl(sysctlArpAnnounce, 2)
  //...
  // 生成masquerade标志用于SNAT规则
    masqueradeValue := 1 << uint(masqueradeBit)
    masqueradeMark := fmt.Sprintf("%#08x/%#08x", masqueradeValue, masqueradeValue)

  // node ip检测
    if nodeIP == nil {
        klog.Warningf("invalid nodeIP, initializing kube-proxy with 127.0.0.1 as nodeIP")
        nodeIP = net.ParseIP("127.0.0.1")
    }

    isIPv6 := utilnet.IsIPv6(nodeIP)
    klog.V(2).Infof("nodeIP: %v, isIPv6: %v", nodeIP, isIPv6)

  // 检测是否有为proxier配置clusterCIDR参数
  // clusterCIDR指定集群中pod使用的网段,以此来区分内部与外部流量
    if len(clusterCIDR) == 0 {
        klog.Warningf("clusterCIDR not specified, unable to distinguish between internal and external traffic")
    } else if utilnet.IsIPv6CIDR(clusterCIDR) != isIPv6 {
        return nil, fmt.Errorf("clusterCIDR %s has incorrect IP version: expect isIPv6=%t", clusterCIDR, isIPv6)
    }

  // 检测是否指定了proxy调度器scheduler算法,如果未指定,则为默认"RR"平均负载算法
    if len(scheduler) == 0 {
        klog.Warningf("IPVS scheduler not specified, use %s by default", DefaultScheduler)
        scheduler = DefaultScheduler
    }

  // healthcheck服务器对象创建
    healthChecker := healthcheck.NewServer(hostname, recorder, nil, nil) 

  // 创建Proxier对象
    proxier := &Proxier{
        //更新SVC、EP信息存放map和changeTracker
    portsMap:              make(map[utilproxy.LocalPort]utilproxy.Closeable),
        serviceMap:            make(proxy.ServiceMap),
        serviceChanges:        proxy.NewServiceChangeTracker(newServiceInfo, &isIPv6, recorder),
        endpointsMap:          make(proxy.EndpointsMap),
        endpointsChanges:      proxy.NewEndpointChangeTracker(hostname, nil, &isIPv6, recorder),
        //同步周期
    syncPeriod:            syncPeriod,
        minSyncPeriod:         minSyncPeriod,

        excludeCIDRs:          excludeCIDRs,
        iptables:              ipt,                      //iptables执行处理器
        masqueradeAll:         masqueradeAll,            //伪装所有访问Service的ClusterIP流量
        masqueradeMark:        masqueradeMark,           //伪装标志号 
        exec:                  exec,                    // osExec命令执行器
        clusterCIDR:           clusterCIDR,
        hostname:              hostname,
        nodeIP:                nodeIP,
        portMapper:            &listenPortOpener{},     
        recorder:              recorder,
        healthChecker:         healthChecker,
        healthzServer:         healthzServer,
        ipvs:                  ipvs,                   //ipvs接口
    ipvsScheduler:         scheduler,              //集群调度算法(默认RR)
        ipGetter:              &realIPGetter{nl: NewNetLinkHandle()}, //node ip获取器
    //iptables规则数据存放buffer
        iptablesData:          bytes.NewBuffer(nil),
        filterChainsData:      bytes.NewBuffer(nil),
        natChains:             bytes.NewBuffer(nil),
        natRules:              bytes.NewBuffer(nil),
        filterChains:          bytes.NewBuffer(nil),
        filterRules:           bytes.NewBuffer(nil),

        netlinkHandle:         NewNetLinkHandle(),         //netlink执行处理器
        ipset:                 ipset,                      //ipset执行处理器
        nodePortAddresses:     nodePortAddresses,    
        networkInterfacer:     utilproxy.RealNetwork{},             
        gracefuldeleteManager: NewGracefulTerminationManager(ipvs),   // RS清理管理器
    }
  // 遍历ipsetInfo定义,初始化kubernetes ipset默认集。(后面在ipset默认集创建时有详细介绍)
    proxier.ipsetList = make(map[string]*IPSet)
    for _, is := range ipsetInfo {
        proxier.ipsetList[is.name] = NewIPSet(ipset, is.name, is.setType, isIPv6, is.comment)
    }
    burstSyncs := 2
    klog.V(3).Infof("minSyncPeriod: %v, syncPeriod: %v, burstSyncs: %d", minSyncPeriod, syncPeriod, burstSyncs)
    proxier.syncRunner = async.NewBoundedFrequencyRunner("sync-runner", proxier.syncProxyRules, minSyncPeriod, syncPeriod, burstSyncs)   //同步runner
  proxier.gracefuldeleteManager.Run()   //后台线程定时(/分钟)清理RS(realServer记录) 
    return proxier, nil
}

3. Proxier 服务与端点更新机制

ipvs模式和iptables模式的service和endpoints更新变化信息同步机制是一致的(更详细说明可参考iptables-mode proxier文章),但为了本文的完整性和相对独立性,这里我们也简单的过一下部分代码。

在构建ipvs-mode proxier对象时指定同步运行器async.NewBoundedFrequencyRunner,同步proxy的规则处理则是syncProxyRules()。同样ipvs-proxier类对象有两个属性对象:serviceChanges(ServiceChangeTracker)和endpointsChanges(EndpointChangeTracker)是就是用来跟踪并记录service和endpoints的变化信息更新至相应的两个属性Items map(serviceChange和endpointsChange)。

pkg/proxy/ipvs/proxier.go:429

proxier.syncRunner = async.NewBoundedFrequencyRunner("sync-runner", proxier.syncProxyRules, minSyncPeriod, syncPeriod, burstSyncs)

在框架层第二层proxy server的运行时最后的调用就是"s.Proxier.SyncLoop()"

pkg/proxy/ipvs/proxier.go:631

func (proxier *Proxier) SyncLoop() {
    // Update healthz timestamp at beginning in case Sync() never succeeds.
  // ...
    proxier.syncRunner.Loop(wait.NeverStop)     //执行NewBoundedFrequencyRunner对象Loop
}

pkg/util/async/bounded_frequency_runner.go:169

func (bfr *BoundedFrequencyRunner) Loop(stop <-chan struct{}) {
    bfr.timer.Reset(bfr.maxInterval)
    for {
        select {
        case <-stop:
            bfr.stop()
            return
        case <-bfr.timer.C():           //定时器方式执行
            bfr.tryRun()
        case <-bfr.run:                 //按需方式执行(发送运行指令信号)
            bfr.tryRun()
        }
    }
}

BoundedFrequencyRunner.tryRun() 按指定频率执行回调函数func "bfr.fn()"

pkg/util/async/bounded_frequency_runner.go:211

func (bfr *BoundedFrequencyRunner) tryRun() {
    bfr.mu.Lock()
    defer bfr.mu.Unlock()

  //限制条件允许运行func
    if bfr.limiter.TryAccept() {
         bfr.fn()                                 // 重点执行部分,调用func,上下文来看此处就是
                                                  // 对syncProxyRules()的调用
        bfr.lastRun = bfr.timer.Now()             // 记录运行时间
        bfr.timer.Stop()                          
        bfr.timer.Reset(bfr.maxInterval)          // 重设下次运行时间
        klog.V(3).Infof("%s: ran, next possible in %v, periodic in %v", bfr.name, bfr.minInterval, bfr.maxInterval)
        return
    }

  //限制条件不允许运行,计算下次运行时间
  elapsed := bfr.timer.Since(bfr.lastRun)    // elapsed:上次运行时间到现在已过多久
  nextPossible := bfr.minInterval - elapsed  // nextPossible:下次运行至少差多久(最小周期)
  nextScheduled := bfr.maxInterval - elapsed // nextScheduled:下次运行最迟差多久(最大周期)
    klog.V(4).Infof("%s: %v since last run, possible in %v, scheduled in %v", bfr.name, elapsed, nextPossible, nextScheduled)

    if nextPossible < nextScheduled {
        bfr.timer.Stop()
        bfr.timer.Reset(nextPossible)
        klog.V(3).Infof("%s: throttled, scheduling run in %v", bfr.name, nextPossible)
    }
}

4. SyncProxyRules 同步 Proxy 规则

syncProxyRules()为proxier的核心逻辑,类似于iptables proxier实现了对apiserver同步的service、endpoints信息的同步与监听,同时在其生成初始和变化时同步ipvs规则(iptables、ipvs虚拟主机、ipset集规则),最终实现kubernetes的"service"机制。

syncProxyRules()代码部分过长,下面将分开对重点部分一一进行分析。

ipvs-mode proxier的同步ipvs规则主要完成以下几个主要步骤操作:

  • 同步与新更service和endpoints;
  • 初始化链和ipset集;
  • 每个服务构建ipvs规则(iptables/ipvs/ipset),服务类型不同生成的规则也相应不同;
  • 清理过旧规则及信息 。

4.1. 更新 service 与 endpoint变化信息

ipvs-mode proxier的service和endpoint变化更新的机制与iptables-mode的完全一致,详细可以参考iptables-mode的"syncProxyRule 同步配置与规则"内的相关内容,这里就不再详细赘述。

Proxier类对象有两个属性:serviceChangesendpointsChanges是就是用来跟踪Service和Endpoint的更新信息,以及两个Tracker及方法:ServiceChangeTracker服务信息变更Tracker,EndpointChangeTracker 端点信息变更Tracker,实时监听apiserver的变更事件。

UpdateServiceMap() svc 服务的更新实现,将serviceChanges的服务项与proxier serviceMap进行更新(合并、删除废弃项)返回,UpdateEndpointsMap() 端点更新的实现,将endpointsChanges的端点项与proxier endpointMap进行更新(合并、删除废弃项)并返回已更新信息。

pkg/proxy/ipvs/proxier.go:730

serviceUpdateResult := proxy.UpdateServiceMap(proxier.serviceMap, proxier.serviceChanges)
endpointUpdateResult := proxy.UpdateEndpointsMap(proxier.endpointsMap, proxier.endpointsChanges)

4.2. 创建 kube 顶层链和连接信息

pkg/proxy/ipvs/proxier.go:748

    proxier.natChains.Reset()     //nat链
    proxier.natRules.Reset()      //nat规则
    proxier.filterChains.Reset()  //filter链
    proxier.filterRules.Reset()   //filter规则
    //写表头
    writeLine(proxier.filterChains, "*filter")
    writeLine(proxier.natChains, "*nat")

    proxier.createAndLinkeKubeChain()  //创建kubernetes的表连接链数据

pkg/proxy/ipvs/proxier.go:1418

func (proxier *Proxier) createAndLinkeKubeChain() {

  //通过iptables-save获取现有的filter和NAT表存在的链数据
    existingFilterChains := proxier.getExistingChains(proxier.filterChainsData, utiliptables.TableFilter)
    existingNATChains := proxier.getExistingChains(proxier.iptablesData, utiliptables.TableNAT)

    // 顶层链数据的构建
  // NAT表链: KUBE-SERVICES / KUBE-POSTROUTING / KUBE-FIREWALL 
  //          KUBE-NODE-PORT / KUBE-LOAD-BALANCER / KUBE-MARK-MASQ
  // Filter表链: KUBE-FORWARD
    for _, ch := range iptablesChains {
    //不存在则创建链,创建顶层链
        if _, err := proxier.iptables.EnsureChain(ch.table, ch.chain); err != nil {
            klog.Errorf("Failed to ensure that %s chain %s exists: %v", ch.table, ch.chain, err)
            return
        }
    //nat表写链
        if ch.table == utiliptables.TableNAT {
            if chain, ok := existingNATChains[ch.chain]; ok {
                writeBytesLine(proxier.natChains, chain)   //现存在的链
            } else {
        // "KUBE-POSTROUTING"
                writeLine(proxier.natChains, utiliptables.MakeChainLine(kubePostroutingChain))
            }
        } else {  // filter表写链
            if chain, ok := existingFilterChains[KubeForwardChain]; ok {
                writeBytesLine(proxier.filterChains, chain)  //现存在的链
            } else {
        // "KUBE-FORWARD"
                writeLine(proxier.filterChains, utiliptables.MakeChainLine(KubeForwardChain))
            }
        }
    }
  // 默认链下创建kubernete服务专用跳转规则
  // iptables -I OUTPUT -t nat --comment "kubernetes service portals" -j KUBE-SERVICES
  // iptables -I PREROUTING -t nat --comment "kubernetes service portals" -j KUBE-SERVICES
  // iptables -I POSTROUTING -t nat --comment "kubernetes postrouting rules" -j KUBE-POSTROUTING
  // iptables -I FORWARD -t filter --comment "kubernetes forwarding rules" -j KUBE-FORWARD
    for _, jc := range iptablesJumpChain {
        args := []string{"-m", "comment", "--comment", jc.comment, "-j", string(jc.to)}
        if _, err := proxier.iptables.EnsureRule(utiliptables.Prepend, jc.table, jc.from, args...); err != nil {
            klog.Errorf("Failed to ensure that %s chain %s jumps to %s: %v", jc.table, jc.from, jc.to, err)
        }
    }
  // 写kubernetes专用的POSTROUTING nat规则
  // -A KUBE-POSTROUTING -m comment --comment "..." -m mark --mark $masqueradeMark -j MASQUERADE
    writeLine(proxier.natRules, []string{
        "-A", string(kubePostroutingChain),
        "-m", "comment", "--comment", `"kubernetes service traffic requiring SNAT"`,
        "-m", "mark", "--mark", proxier.masqueradeMark,
        "-j", "MASQUERADE",
    }...)

  // 写kubernetes专用的masquerade伪装地址标记规则
  // -A KUBE-MARK-MASQ -j MARK --set-xmark $masqueradeMark
    writeLine(proxier.natRules, []string{
        "-A", string(KubeMarkMasqChain),
        "-j", "MARK", "--set-xmark", proxier.masqueradeMark,
    }...)
}

4.3. Dummy 接口和 ipset 默认集创建

pkg/proxy/ipvs/proxier.go:760

 //为服务地址的绑定,确保已创建虚拟接口kube-ipvs0
    _, err := proxier.netlinkHandle.EnsureDummyDevice(DefaultDummyDevice)
    if err != nil {
        klog.Errorf("Failed to create dummy interface: %s, error: %v", DefaultDummyDevice, err)
        return
    }

 // 确保kubernetes专用的ipset集已创建
    for _, set := range proxier.ipsetList {
        if err := ensureIPSet(set); err != nil {
            return
        }
        set.resetEntries()
    }

proxier.ipsetList的定义信息,在proxier对象创建时初始化了ipsetList列表

pkg/proxy/ipvs/proxier.go:113

var ipsetInfo = []struct {
    name    string             //ipset set名称
  setType utilipset.Type     //set类型{HashIPPortIP|HashIPPort|HashIPPortNet|BitmapPort}
    comment string             //comment描述信息
}{
    {kubeLoopBackIPSet, utilipset.HashIPPortIP, kubeLoopBackIPSetComment},
   //...
}
ipset集名称 类型 描述
KUBE-LOOP-BACK hash:ip,port,ip Kubernetes endpoints dst ip:port, source ip for solving hairpin purpose
KUBE-CLUSTER-IP hash:ip,port Kubernetes service cluster ip + port for masquerade purpose
KUBE-EXTERNAL-IP hash:ip,port Kubernetes service external ip + port for masquerade and filter purpose
KUBE-LOAD-BALANCER hash:ip,port Kubernetes service lb portal
KUBE-LOAD-BALANCER-FW hash:ip,port Kubernetes service load balancer ip + port for load balancer with sourceRange
KUBE-LOAD-BALANCER-LOCAL hash:ip,port Kubernetes service load balancer ip + port with externalTrafficPolicy=local
KUBE-LOAD-BALANCER-SOURCE-IP hash:ip,port,ip Kubernetes service load balancer ip + port + source IP for packet filter purpose
KUBE-LOAD-BALANCER-SOURCE-CIDR hash:ip,port,net Kubernetes service load balancer ip + port + source cidr for packet filter purpose
KUBE-NODE-PORT-TCP BitmapPort Kubernetes nodeport TCP port for masquerade purpose
KUBE-NODE-PORT-LOCAL-TCP BitmapPort BitmapPort,Kubernetes nodeport TCP port with externalTrafficPolicy=local
KUBE-NODE-PORT-UDP BitmapPort Kubernetes nodeport UDP port for masquerade purpose
KUBE-NODE-PORT-LOCAL-UDP BitmapPort Kubernetes nodeport UDP port with externalTrafficPolicy=local
KUBE-NODE-PORT-SCTP BitmapPort Kubernetes nodeport SCTP port for masquerade purpose
KUBE-NODE-PORT-LOCAL-SCTP BitmapPort Kubernetes nodeport SCTP port with externalTrafficPolicy=local

4.4. 每个服务生成 ipvs 规则

代码逻辑包含在一个for循环内,对serviceMap内的每个服务进行遍历处理,对不同的服务类型(clusterip/nodePort/externalIPs/load-balancer)进行不同的处理(ipset集/ipvs虚拟主机/ipvs后端服务器)。

ipvs模式,通过svc创建的集群都绑定在默认dummy(kube-ipvs0)虚拟网卡,创建ipvs集群IP执行以下几项操作:

  • 节点中存在虚拟接口为 kube-ipvs0,且服务 IP 地址绑定到虚拟接口
  • 分别为每个kube服务 IP 地址创建 IPVS 虚拟服务器
  • 为每个 IPVS 虚拟服务器创建RealServers (kube服务 endpoints)

pkg/proxy/ipvs/proxier.go:784

for svcName, svc := range proxier.serviceMap {
   //...... 后面详细分析
}

基于此服务的有效endpoint列表,更新KUBE-LOOP-BACK的ipset集,以备后面生成相应iptables规则(SNAT伪装地址)。

pkg/proxy/ipvs/proxier.go:796

        for _, e := range proxier.endpointsMap[svcName] {
            ep, ok := e.(*proxy.BaseEndpointInfo)
            if !ok {
                klog.Errorf("Failed to cast BaseEndpointInfo %q", e.String())
                continue
            }
            if !ep.IsLocal {                //非本地
                continue
            }
            epIP := ep.IP()                 //端点IP
            epPort, err := ep.Port()        //端点Port

            if epIP == "" || err != nil {   //有效IP和端口正常
                continue
            }
      // 构造ipset集的entry记录项
            entry := &utilipset.Entry{
                IP:       epIP,
                Port:     epPort,
                Protocol: protocol,
                IP2:      epIP,
                SetType:  utilipset.HashIPPortIP,
            }
      // 类型校验KUBE-LOOP-BACK集合entry记录项
            if valid := proxier.ipsetList[kubeLoopBackIPSet].validateEntry(entry); !valid {
                klog.Errorf("%s", fmt.Sprintf(EntryInvalidErr, entry, proxier.ipsetList[kubeLoopBackIPSet].Name))
                continue
            }
      // 插入此entry记录至active记录队列
            proxier.ipsetList[kubeLoopBackIPSet].activeEntries.Insert(entry.String())
        }

clusterIP服务类型流量的承接(clusterIP为默认方式,仅资源集群内可访问),ipset集KUBE-CLUSTER-IP更新,以备后面生成相应iptables规则。

pkg/proxy/ipvs/proxier.go:827

        //构建ipset entry
        entry := &utilipset.Entry{                      
            IP:       svcInfo.ClusterIP.String(),
            Port:     svcInfo.Port,
            Protocol: protocol,
            SetType:  utilipset.HashIPPort,
        }
        // 类型校验ipset entry
        if valid := proxier.ipsetList[kubeClusterIPSet].validateEntry(entry); !valid {
            klog.Errorf("%s", fmt.Sprintf(EntryInvalidErr, entry, proxier.ipsetList[kubeClusterIPSet].Name))
            continue
        }
    // 名为KUBE-CLUSTER-IP的ipset集插入entry,以备后面统一生成IPtables规则
        proxier.ipsetList[kubeClusterIPSet].activeEntries.Insert(entry.String())
        // 构建ipvs虚拟服务器VS服务对象
        serv := &utilipvs.VirtualServer{
            Address:   svcInfo.ClusterIP,
            Port:      uint16(svcInfo.Port),
            Protocol:  string(svcInfo.Protocol),
            Scheduler: proxier.ipvsScheduler,
        }
    // 设置IPVS服务的会话保持标志和超时时间
        if svcInfo.SessionAffinityType == v1.ServiceAffinityClientIP {
            serv.Flags |= utilipvs.FlagPersistent
            serv.Timeout = uint32(svcInfo.StickyMaxAgeSeconds)
        }

    // 将clusterIP绑定至dummy虚拟接口上,syncService()处理中需置bindAddr地址为True。
    // ipvs为服务创建VS(虚拟主机)
        if err := proxier.syncService(svcNameString, serv, true); err == nil {
            activeIPVSServices[serv.String()] = true
            activeBindAddrs[serv.Address.String()] = true
      // 为虚拟主机/服务(vip)同步endpoints信息。
      // IPVS为VS更新RS(realServer后端服务器)
            if err := proxier.syncEndpoint(svcName, false, serv); err != nil {
                klog.Errorf("Failed to sync endpoint for service: %v, err: %v", serv, err)
            }
        } else {
            klog.Errorf("Failed to sync service: %v, err: %v", serv, err)
        }

syncService() 更新和同步ipvs服务信息及服务IP与dummy接口的绑定

pkg/proxy/ipvs/proxier.go:1498

func (proxier *Proxier) syncService(svcName string, vs *utilipvs.VirtualServer, bindAddr bool) error {
  //获取IPVS虚拟主机服务信息
    appliedVirtualServer, _ := proxier.ipvs.GetVirtualServer(vs) 
  //无此虚拟主机服务或此服务信息变更
    if appliedVirtualServer == nil || !appliedVirtualServer.Equal(vs) {
        if appliedVirtualServer == nil {
      // 服务未找到,则创建新的服务
            klog.V(3).Infof("Adding new service %q %s:%d/%s", svcName, vs.Address, vs.Port, vs.Protocol)
            if err := proxier.ipvs.AddVirtualServer(vs); err != nil {
                klog.Errorf("Failed to add IPVS service %q: %v", svcName, err)
                return err
            }
        } else {
      // 服务信息改变,则更新存在服务信息,在更新期间服务VIP不会关闭
            klog.V(3).Infof("IPVS service %s was changed", svcName)
            if err := proxier.ipvs.UpdateVirtualServer(vs); err != nil {
                klog.Errorf("Failed to update IPVS service, err:%v", err)
                return err
            }
        }

  // 将服务IP绑定到dummy接口上
    if bindAddr {
        klog.V(4).Infof("Bind addr %s", vs.Address.String())
        _, err := proxier.netlinkHandle.EnsureAddressBind(vs.Address.String(), DefaultDummyDevice) //netlinkHandle处理的实现在文章最后的netlink工具介绍部分详细说明 
        if err != nil {
            klog.Errorf("Failed to bind service address to dummy device %q: %v", svcName, err)
            return err
        }
    }
    return nil
}

syncEndpoint() 为虚拟主机/服务(clusterip)同步endpoints信息,实现ipvs为VS更新RS(realServer后端服务器)。

pkg/proxy/ipvs/proxier.go:1532

func (proxier *Proxier) syncEndpoint(svcPortName proxy.ServicePortName, onlyNodeLocalEndpoints bool, vs *utilipvs.VirtualServer) error {
    appliedVirtualServer, err := proxier.ipvs.GetVirtualServer(vs)
    if err != nil || appliedVirtualServer == nil {
        klog.Errorf("Failed to get IPVS service, error: %v", err)
        return err
    }

  // curEndpoints表示当前系统IPVS目标列表
    curEndpoints := sets.NewString()
  // newEndpoints表示从apiServer监听到的Endpoints
    newEndpoints := sets.NewString()

  // 依据虚拟服务器获取RS(realservers)列表
    curDests, err := proxier.ipvs.GetRealServers(appliedVirtualServer)
    if err != nil {
        klog.Errorf("Failed to list IPVS destinations, error: %v", err)
        return err
    }
    for _, des := range curDests {
        curEndpoints.Insert(des.String())    // 写入curEndpoints
    }

  //迭代endpointsMaps信息,将非本地的enpoints写入newEndpoints
    for _, epInfo := range proxier.endpointsMap[svcPortName] {
        if onlyNodeLocalEndpoints && !epInfo.GetIsLocal() {
            continue
        }
        newEndpoints.Insert(epInfo.String())   
    }

    // 创建新的endpoints
    for _, ep := range newEndpoints.List() {
        ip, port, err := net.SplitHostPort(ep)
        if err != nil {
            klog.Errorf("Failed to parse endpoint: %v, error: %v", ep, err)
            continue
        }
        portNum, err := strconv.Atoi(port)
        if err != nil {
            klog.Errorf("Failed to parse endpoint port %s, error: %v", port, err)
            continue
        }

        newDest := &utilipvs.RealServer{
            Address: net.ParseIP(ip),
            Port:    uint16(portNum),
            Weight:  1,
        }
    //判断当前系统ipvs列表是否存在
        if curEndpoints.Has(ep) {
      //检测是否在gracefulDelete列表,如果是则此处立即删除
            uniqueRS := GetUniqueRSName(vs, newDest)
            if !proxier.gracefuldeleteManager.InTerminationList(uniqueRS) {
                continue
            }
            klog.V(5).Infof("new ep %q is in graceful delete list", uniqueRS)
            err := proxier.gracefuldeleteManager.MoveRSOutofGracefulDeleteList(uniqueRS)
            if err != nil {
                klog.Errorf("Failed to delete endpoint: %v in gracefulDeleteQueue, error: %v", ep, err)
                continue
            }
        }
    // 不存在则新增RealServer项(对应目标endpoint)
        err = proxier.ipvs.AddRealServer(appliedVirtualServer, newDest)
        if err != nil {
            klog.Errorf("Failed to add destination: %v, error: %v", newDest, err)
            continue
        }
    }
  // 删除过旧的endpoints
    for _, ep := range curEndpoints.Difference(newEndpoints).UnsortedList() {
    // 如果curEndpoint在gracefulDelete内,跳过
        uniqueRS := vs.String() + "/" + ep
        if proxier.gracefuldeleteManager.InTerminationList(uniqueRS) {
            continue
        }
        ip, port, err := net.SplitHostPort(ep)
        if err != nil {
            klog.Errorf("Failed to parse endpoint: %v, error: %v", ep, err)
            continue
        }
        portNum, err := strconv.Atoi(port)
        if err != nil {
            klog.Errorf("Failed to parse endpoint port %s, error: %v", port, err)
            continue
        }

        delDest := &utilipvs.RealServer{
            Address: net.ParseIP(ip),
            Port:    uint16(portNum),
        }

        klog.V(5).Infof("Using graceful delete to delete: %v", uniqueRS)
    // 删除RS
        err = proxier.gracefuldeleteManager.GracefulDeleteRS(appliedVirtualServer, delDest)
        if err != nil {
            klog.Errorf("Failed to delete destination: %v, error: %v", uniqueRS, err)
            continue
        }
    }
    return nil
}

externalIPs服务类型流量的承接,服务是否启用ExternalIPs,在指定的Node上开启监听端口(代码逻辑判断是否为本地ip),而非像nodeport所有节点监听。ipset集KUBE-EXTERNAL-IP更新,以备后面生成相应iptables规则。

pkg/proxy/ipvs/proxier.go:866

        for _, externalIP := range svcInfo.ExternalIPs {
            if local, err := utilproxy.IsLocalIP(externalIP); err != nil {
                klog.Errorf("can't determine if IP is local, assuming not: %v", err)
        // 如果指定的externealIP为本地地址且协议不为SCTP
            } else if local && (svcInfo.GetProtocol() != v1.ProtocolSCTP) {
                lp := utilproxy.LocalPort{
                    Description: "externalIP for " + svcNameString,
                    IP:          externalIP,
                    Port:        svcInfo.Port,
                    Protocol:    protocol,
                }
                if proxier.portsMap[lp] != nil {   //端口已存在
                    klog.V(4).Infof("Port %s was open before and is still needed", lp.String())
                    replacementPortsMap[lp] = proxier.portsMap[lp]
                } else {
                    socket, err := proxier.portMapper.OpenLocalPort(&lp)  //打开本地端口socket
                    if err != nil {
                        msg := fmt.Sprintf("can't open %s, skipping this externalIP: %v", lp.String(), err)

                        proxier.recorder.Eventf(       //通知事件
                            &v1.ObjectReference{
                                Kind:      "Node",
                                Name:      proxier.hostname,
                                UID:       types.UID(proxier.hostname),
                                Namespace: "",
                            }, v1.EventTypeWarning, err.Error(), msg)
                        klog.Error(msg)
                        continue
                    }
                    replacementPortsMap[lp] = socket    //存放端口信息
                }
            } 

            // 创建ipset entry
            entry := &utilipset.Entry{
                IP:       externalIP,
                Port:     svcInfo.Port,
                Protocol: protocol,
                SetType:  utilipset.HashIPPort,
            }
            // We have to SNAT packets to external IPs.
            if valid := proxier.ipsetList[kubeExternalIPSet].validateEntry(entry); !valid {
                klog.Errorf("%s", fmt.Sprintf(EntryInvalidErr, entry, proxier.ipsetList[kubeExternalIPSet].Name))
                continue
            }
      // 名为KUBE-EXTERNAL-IP的ipset集插入entry,以备后面统一生成IPtables规则
            proxier.ipsetList[kubeExternalIPSet].activeEntries.Insert(entry.String())

            // 为服务定义ipvs虚拟主机信息
            serv := &utilipvs.VirtualServer{
                Address:   net.ParseIP(externalIP),
                Port:      uint16(svcInfo.Port),
                Protocol:  string(svcInfo.Protocol),
                Scheduler: proxier.ipvsScheduler,
            }
            if svcInfo.SessionAffinityType == v1.ServiceAffinityClientIP {
                serv.Flags |= utilipvs.FlagPersistent
                serv.Timeout = uint32(svcInfo.StickyMaxAgeSeconds)
            }
      // 将clusterIP绑定至dummy虚拟接口上,syncService()处理中需置bindAddr地址为True。
      // ipvs为服务创建VS(虚拟主机)
      // 为虚拟主机/服务同步endpoints信息。
      // IPVS为VS更新RS(realServer后端服务器)
      //...(同clusterip)
        }

load-balancer服务类型流量的承接,服务的LoadBalancerSourceRanges和externalTrafficPolicy=local被指定时将对KUBE-LOAD-BALANCER-LOCAL、KUBE-LOAD-BALANCER-FW、KUBE-LOAD-BALANCER-SOURCE-CIDR、KUBE-LOAD-BALANCER-SOURCE-IP ipset集更新,以备后面生成相应iptables规则。

pkg/proxy/ipvs/proxier.go:937

        for _, ingress := range svcInfo.LoadBalancerStatus.Ingress {
            if ingress.IP != "" {
                // 构建ipset entry 
                entry = &utilipset.Entry{
                    IP:       ingress.IP,
                    Port:     svcInfo.Port,
                    Protocol: protocol,
                    SetType:  utilipset.HashIPPort,
                }
        // 增加SLB(service load balancer)ingressIP:Port与kube服务IP集对应
        // KUBE-LOAD-BALANCER ipset集更新 
                if valid := proxier.ipsetList[kubeLoadBalancerSet].validateEntry(entry); !valid {
                    klog.Errorf("%s", fmt.Sprintf(EntryInvalidErr, entry, proxier.ipsetList[kubeLoadBalancerSet].Name))
                    continue
                }
                proxier.ipsetList[kubeLoadBalancerSet].activeEntries.Insert(entry.String())
        // 服务指定externalTrafficPolicy=local时,KUBE-LOAD-BALANCER-LOCAL ipset集更新
                if svcInfo.OnlyNodeLocalEndpoints {
                    if valid := proxier.ipsetList[kubeLoadBalancerLocalSet].validateEntry(entry); !valid {
                        klog.Errorf("%s", fmt.Sprintf(EntryInvalidErr, entry, proxier.ipsetList[kubeLoadBalancerLocalSet].Name))
                        continue
                    }
                    proxier.ipsetList[kubeLoadBalancerLocalSet].activeEntries.Insert(entry.String())
                }
        // 服务的LoadBalancerSourceRanges被指定时,基于源IP保护的防火墙策略开启,KUBE-LOAD-BALANCER-FW ipset集更新
                if len(svcInfo.LoadBalancerSourceRanges) != 0 {
                    if valid := proxier.ipsetList[kubeLoadbalancerFWSet].validateEntry(entry); !valid {
                        klog.Errorf("%s", fmt.Sprintf(EntryInvalidErr, entry, proxier.ipsetList[kubeLoadbalancerFWSet].Name))
                        continue
                    }
                    proxier.ipsetList[kubeLoadbalancerFWSet].activeEntries.Insert(entry.String())
                    allowFromNode := false
                    for _, src := range svcInfo.LoadBalancerSourceRanges {
                        // 构建ipset entry
                        entry = &utilipset.Entry{
                            IP:       ingress.IP,
                            Port:     svcInfo.Port,
                            Protocol: protocol,
                            Net:      src,
                            SetType:  utilipset.HashIPPortNet,
                        }
            // 枚举所有源CIDR白名单列表,KUBE-LOAD-BALANCER-SOURCE-CIDR ipset集更新
                        if valid := proxier.ipsetList[kubeLoadBalancerSourceCIDRSet].validateEntry(entry); !valid {
                            klog.Errorf("%s", fmt.Sprintf(EntryInvalidErr, entry, proxier.ipsetList[kubeLoadBalancerSourceCIDRSet].Name))
                            continue
                        }
                        proxier.ipsetList[kubeLoadBalancerSourceCIDRSet].activeEntries.Insert(entry.String())

                        // ignore error because it has been validated
                        _, cidr, _ := net.ParseCIDR(src)
                        if cidr.Contains(proxier.nodeIP) {
                            allowFromNode = true
                        }
                    }

          // 允许来自Node流量(LB对应后端hosts之间交互)
                    if allowFromNode {
                        entry = &utilipset.Entry{
                            IP:       ingress.IP,
                            Port:     svcInfo.Port,
                            Protocol: protocol,
                            IP2:      ingress.IP,
                            SetType:  utilipset.HashIPPortIP,
                        }

            // 枚举所有白名单源IP列表,KUBE-LOAD-BALANCER-SOURCE-IP ipset集更新
                        if valid := proxier.ipsetList[kubeLoadBalancerSourceIPSet].validateEntry(entry); !valid {
                            klog.Errorf("%s", fmt.Sprintf(EntryInvalidErr, entry, proxier.ipsetList[kubeLoadBalancerSourceIPSet].Name))
                            continue
                        }
                        proxier.ipsetList[kubeLoadBalancerSourceIPSet].activeEntries.Insert(entry.String())
                    }
                }

                // 构建ipvs 虚拟主机对象
                serv := &utilipvs.VirtualServer{
                    Address:   net.ParseIP(ingress.IP),   // SLB ip
                    Port:      uint16(svcInfo.Port),      // SLB 端口
                    Protocol:  string(svcInfo.Protocol),  // 协议
                    Scheduler: proxier.ipvsScheduler,     // RR
                }
                if svcInfo.SessionAffinityType == v1.ServiceAffinityClientIP {
                    serv.Flags |= utilipvs.FlagPersistent
                    serv.Timeout = uint32(svcInfo.StickyMaxAgeSeconds)
                }
         // ipvs为服务创建VS(虚拟主机),LB ingressIP绑定dummy接口
         // ipvs为VS更新RS(realServer后端服务器)
               //...(同clusterip)
            }
        }

NodePort服务类型流量的承接,服务将在每个节点上都将开启指定的nodeport端口,并更新相应的ipset集。

pkg/proxy/ipvs/proxier.go:1040

if svcInfo.NodePort != 0 {
            addresses, err := utilproxy.GetNodeAddresses(proxier.nodePortAddresses, proxier.networkInterfacer)  // 获取node addresses
            if err != nil {
                klog.Errorf("Failed to get node ip address matching nodeport cidr: %v", err)
                continue
            }

            var lps []utilproxy.LocalPort
            for address := range addresses {   
                lp := utilproxy.LocalPort{
                    Description: "nodePort for " + svcNameString,
                    IP:          address,
                    Port:        svcInfo.NodePort,
                    Protocol:    protocol,
                }
                if utilproxy.IsZeroCIDR(address) {
                    // Empty IP address means all
                    lp.IP = ""
                    lps = append(lps, lp)       
                    break
                }
        lps = append(lps, lp)  //整理与格式化后的lps列表
            }

      // 为node节点的IPs打开端口并保存持有socket句柄
            for _, lp := range lps {
                if proxier.portsMap[lp] != nil {
                    klog.V(4).Infof("Port %s was open before and is still needed", lp.String())
                    replacementPortsMap[lp] = proxier.portsMap[lp]

                } else if svcInfo.GetProtocol() != v1.ProtocolSCTP {
          // 打开和监听端口(非SCTP协议)
                    socket, err := proxier.portMapper.OpenLocalPort(&lp)
                    if err != nil {
                        klog.Errorf("can't open %s, skipping this nodePort: %v", lp.String(), err)
                        continue
                    }
                    if lp.Protocol == "udp" {
            // UDP协议,清理udp conntrack记录
                        isIPv6 := utilnet.IsIPv6(svcInfo.ClusterIP)
                        conntrack.ClearEntriesForPort(proxier.exec, lp.Port, isIPv6, v1.ProtocolUDP)
                    }
                    replacementPortsMap[lp] = socket
                } //socket保存
            }

            // Nodeports无论是否为本地都需要SNAT
            // 构建ipset entry
            entry = &utilipset.Entry{
                // No need to provide ip info
                Port:     svcInfo.NodePort,      
                Protocol: protocol,
                SetType:  utilipset.BitmapPort,
            }
            var nodePortSet *IPSet
      //基于协议类型选择ipset集
            switch protocol {      
            case "tcp":               // KUBE-NODE-PORT-TCP 
                nodePortSet = proxier.ipsetList[kubeNodePortSetTCP]
            case "udp":               // KUBE-NODE-PORT-UDP
                nodePortSet = proxier.ipsetList[kubeNodePortSetUDP]
            case "sctp":              // KUBE-NODE-PORT-SCTP
                nodePortSet = proxier.ipsetList[kubeNodePortSetSCTP]
            default:
                klog.Errorf("Unsupported protocol type: %s", protocol)
            }
            if nodePortSet != nil {
                if valid := nodePortSet.validateEntry(entry); !valid {
                    klog.Errorf("%s", fmt.Sprintf(EntryInvalidErr, entry, nodePortSet.Name))
                    continue
                }
        // 更新ipset集
                nodePortSet.activeEntries.Insert(entry.String())
            }

      // 服务externaltrafficpolicy=local指定时,基于协议类型更新ipset集entry
            if svcInfo.OnlyNodeLocalEndpoints {
                var nodePortLocalSet *IPSet
                switch protocol {
                case "tcp":           //KUBE-NODE-PORT-LOCAL-TCP
                    nodePortLocalSet = proxier.ipsetList[kubeNodePortLocalSetTCP]
                case "udp":           //KUBE-NODE-PORT-LOCAL-UDP
                    nodePortLocalSet = proxier.ipsetList[kubeNodePortLocalSetUDP]
                case "sctp":          //KUBE-NODE-PORT-LOCAL-SCTP
                    nodePortLocalSet = proxier.ipsetList[kubeNodePortLocalSetSCTP]
                default:
                    klog.Errorf("Unsupported protocol type: %s", protocol)
                }
                if nodePortLocalSet != nil {
                    if valid := nodePortLocalSet.validateEntry(entry); !valid {
                        klog.Errorf("%s", fmt.Sprintf(EntryInvalidErr, entry, nodePortLocalSet.Name))
                        continue
                    }
          //  更新ipset集
                    nodePortLocalSet.activeEntries.Insert(entry.String())
                }
            }

      // 为Node每个ip address创建ipvs路由项(VS/RS)
            var nodeIPs []net.IP
            for address := range addresses {
                if !utilproxy.IsZeroCIDR(address) {
                    nodeIPs = append(nodeIPs, net.ParseIP(address))
                    continue
                }
                // zero cidr
                nodeIPs, err = proxier.ipGetter.NodeIPs()  
                if err != nil {
                    klog.Errorf("Failed to list all node IPs from host, err: %v", err)
                }
            }
            for _, nodeIP := range nodeIPs {
                // 构建ipvs VS对象
                serv := &utilipvs.VirtualServer{
                    Address:   nodeIP,                     //node ip地址
                    Port:      uint16(svcInfo.NodePort),   //node端口
                    Protocol:  string(svcInfo.Protocol),   //协议
                    Scheduler: proxier.ipvsScheduler,      //RR
                }
                if svcInfo.SessionAffinityType == v1.ServiceAffinityClientIP {
                    serv.Flags |= utilipvs.FlagPersistent
                    serv.Timeout = uint32(svcInfo.StickyMaxAgeSeconds)
                }
        // 这里不需要将Node IP绑定到dummy接口,参数值为false
        // ipvs为服务创建VS(虚拟主机)
        // ipvs为VS更新RS(realServer后端服务器)
        //...(同clusterip)
        }

4.5. SyncIPSetEntries 同步 ipset 记录

pkg/proxy/ipvs/proxier.go:1176

for _, set := range proxier.ipsetList {
    set.syncIPSetEntries()
}

pkg/proxy/ipvs/ipset.go:125

func (set *IPSet) syncIPSetEntries() {
    appliedEntries, err := set.handle.ListEntries(set.Name)
    if err != nil {
        klog.Errorf("Failed to list ip set entries, error: %v", err)
        return
    }

  // currentIPSetEntries代表从apiServer上一直监听着的endpoints列表
    currentIPSetEntries := sets.NewString()
    for _, appliedEntry := range appliedEntries {
        currentIPSetEntries.Insert(appliedEntry)
    }

  // 求差集
  // s1 = {a1, a2, a3}
  // s2 = {a1, a2, a4, a5}
  // s1.Difference(s2) = {a3}
  // s2.Difference(s1) = {a4,a5}
    if !set.activeEntries.Equal(currentIPSetEntries) {
        // 清理过旧记录(取currentIPSetEntries在activeEntries中没有的entries)
        for _, entry := range currentIPSetEntries.Difference(set.activeEntries).List() {
            if err := set.handle.DelEntry(entry, set.Name); err != nil {
                if !utilipset.IsNotFoundError(err) {
                    klog.Errorf("Failed to delete ip set entry: %s from ip set: %s, error: %v", entry, set.Name, err)
                }
            } else {
                klog.V(3).Infof("Successfully delete legacy ip set entry: %s from ip set: %s", entry, set.Name)
            }
        }
        // 新增记录(取activeEntries在currentIPSetEntries中没有的entries)
        for _, entry := range set.activeEntries.Difference(currentIPSetEntries).List() {
            if err := set.handle.AddEntry(entry, &set.IPSet, true); err != nil {
                klog.Errorf("Failed to add entry: %v to ip set: %s, error: %v", entry, set.Name, err)
            } else {
                klog.V(3).Infof("Successfully add entry: %v to ip set: %s", entry, set.Name)
            }
        }
    }
}

4.6. 创建 iptables 规则数据

pkg/proxy/ipvs/proxier.go:1182

proxier.writeIptablesRules()

基于ipset定义创建iptables NAT表的kubernetes初始固定链规则数据。

pkg/proxy/ipvs/proxier.go:1269

    for _, set := range ipsetWithIptablesChain {
        if _, find := proxier.ipsetList[set.name]; find && !proxier.ipsetList[set.name].isEmpty() {
            args = append(args[:0], "-A", set.from)
            if set.protocolMatch != "" {
                args = append(args, "-p", set.protocolMatch)
            }
            args = append(args,
                "-m", "comment", "--comment", proxier.ipsetList[set.name].getComment(),
                "-m", "set", "--match-set", set.name,
                set.matchType,
            )
      // -A $setFrom -p $prot -m comment --comment $commentStr 
      //    -m set --match-set $setName $setType -j $setTo
            writeLine(proxier.natRules, append(args, "-j", set.to)...)
        }
    }

依据ipsetWithIptablesChain定义生成以下创建固定链规则数据

KUBE-POSTROUTING匹配KUBE-LOOP-BACK ipset表则伪装地址

-A KUBE-POSTROUTING -m comment --comment "Kubernetes endpoints dst ip:port, source ip for solving hairpin purpose" -m set --match-set KUBE-LOOP-BACK dst,dst,src -j MASQUERADE

LoadBalancer服务类型相关规则

-A KUBE-SERVICES -m comment --comment "Kubernetes service lb portal" -m set --match-set KUBE-LOAD-BALANCER dst,dst -j KUBE-LOAD-BALANCER

-A KUBE-LOAD-BALANCER -m comment --comment "Kubernetes service load balancer ip + port for load balancer with sourceRange" -m set --match-set KUBE-LOAD-BALANCER-FW dst,dst -j KUBE-FIREWALL -A KUBE-FIREWALL -m comment --comment "Kubernetes service load balancer ip + port + source cidr for packet filter" -m set --match-set KUBE-LOAD-BALANCER-SOURCE-CIDR dst,dst,src -j RETURN -A KUBE-FIREWALL -m comment --comment "Kubernetes service load balancer ip + port + source IP for packet filter purpose" -m set --match-set KUBE-LOAD-BALANCER-SOURCE-IP dst,dst,src -j RETURN -A KUBE-LOAD-BALANCER -m comment --comment "Kubernetes service load balancer ip + port with externalTrafficPolicy=local" -m set --match-set KUBE-LOAD-BALANCER-LOCAL dst,dst -j RETURN

Nodeport服务类型相关规则

-A KUBE-NODE-PORT -p tcp -m comment --comment "Kubernetes service load balancer ip + port with externalTrafficPolicy=local" -m set --match-set KUBE-NODE-PORT-LOCAL-TCP dst -j RETURN -A KUBE-NODE-PORT -p tcp -m comment --comment "Kubernetes nodeport TCP port for masquerade purpose" -m set --match-set KUBE-NODE-PORT-TCP dst -j KUBE-MARK-MASQ -A KUBE-NODE-PORT -p udp -m comment --comment "Kubernetes nodeport UDP port with externalTrafficPolicy=local" -m set --match-set KUBE-NODE-PORT-LOCAL-UDP dst -j RETURN -A KUBE-NODE-PORT -p udp -m comment --comment "Kubernetes nodeport UDP port for masquerade purpose" -m set --match-set KUBE-NODE-PORT-UDP dst -j KUBE-MARK-MASQ -A KUBE-SERVICES -p sctp -m comment --comment "Kubernetes nodeport SCTP port for masquerade purpose" -m set --match-set KUBE-NODE-PORT-SCTP dst -j KUBE-NODE-PORT -A KUBE-NODE-PORT -p sctp -m comment --comment "Kubernetes nodeport SCTP port with externalTrafficPolicy=local" -m set --match-set KUBE-NODE-PORT-LOCAL-SCTP dst -j RETURN

kube-proxy启动参数"--masquerade-all=true", 针对类型为clusterip服务生成相应的NAT表KUBE-SERVICES链规则数据,masquerade-all实现访问service ip流量伪装。

pkg/proxy/ipvs/proxier.go:1284

//ipset名称为"KUBE-CLUSTER-IP"不为空,即clusterip类型服务    
if !proxier.ipsetList[kubeClusterIPSet].isEmpty() {
        args = append(args[:0],
            "-A", string(kubeServicesChain),
            "-m", "comment", "--comment", proxier.ipsetList[kubeClusterIPSet].getComment(),
            "-m", "set", "--match-set", kubeClusterIPSet,
        )
     //当proxy配置为masqueradeAll=true
        if proxier.masqueradeAll {
      //nat表:-A KUBE-SERVICES -m comment --comment "Kubernetes service cluster ip + port for masquerade purpose" -m set --match-set KUBE-CLUSTER-IP dst,dst -j KUBE-MARK-MASQ
            writeLine(proxier.natRules, append(args, "dst,dst", "-j", string(KubeMarkMasqChain))...)
        } else if len(proxier.clusterCIDR) > 0 {
      //当指定了clusterCIDR,针对非集群到服务VIP的流量masquerades规则 (dst,dst 目标ip:目标端口)
      // nat表:-A KUBE-SERVICES -m comment --comment "Kubernetes service cluster ip + port for masquerade purpose" -m set --match-set KUBE-CLUSTER-IP dst,dst ! -s $clusterCIDR -j KUBE-MARK-MASQ
            writeLine(proxier.natRules, append(args, "dst,dst", "! -s", proxier.clusterCIDR, "-j", string(KubeMarkMasqChain))...)
        } else {
      // 所有来自服务VIP出流量masquerades规则 (src,dst 源ip:目标端口)
      // 如:VIP:<random port> to VIP:<service port>
      // nat表:-A KUBE-SERVICES -m comment --comment "Kubernetes service cluster ip + port for masquerade purpose" -m set --match-set KUBE-CLUSTER-IP src,dst -j KUBE-MARK-MASQ
            writeLine(proxier.natRules, append(args, "src,dst", "-j", string(KubeMarkMasqChain))...)
        }
    }

为服务externalIPs专用ipset集(存在配置externalIPs的服务)生成相应的iptables NAT表规则数据。

pkg/proxy/ipvs/proxier.go:1311

    if !proxier.ipsetList[kubeExternalIPSet].isEmpty() {
    // 为external IPs添加masquerade规则
        args = append(args[:0],
            "-A", string(kubeServicesChain),
            "-m", "comment", "--comment", proxier.ipsetList[kubeExternalIPSet].getComment(),
            "-m", "set", "--match-set", kubeExternalIPSet,
            "dst,dst",
        )
    // -A KUBE-SERVICES -m comment --comment "Kubernetes service external ip + port for masquerade and filter purpose" -m set --match-set KUBE-EXTERNAL-IP dst,dst -j KUBE-MARK-MASQ
        writeLine(proxier.natRules, append(args, "-j", string(KubeMarkMasqChain))...)

    // 允许external ips流量,而非来自本地网桥流量(如来自一个容器流量或本地处理的forward至服务流量)
        externalTrafficOnlyArgs := append(args,
            "-m", "physdev", "!", "--physdev-is-in",
            "-m", "addrtype", "!", "--src-type", "LOCAL")
    // -m set match-set KUBE-EXTERNAL-IP dst,dst -m PHYSDEV  ! --physdev-is-in -m addrtype ! --src-type LOCAL -j ACCEPT
        writeLine(proxier.natRules, append(externalTrafficOnlyArgs, "-j", "ACCEPT")...)
        dstLocalOnlyArgs := append(args, "-m", "addrtype", "--dst-type", "LOCAL")

    // 识别与允许本地流量
    // -m set match-set KUBE-EXTERNAL-IP dst,dst -m addrtype --dst-type LOCAL -j ACCEPT
        writeLine(proxier.natRules, append(dstLocalOnlyArgs, "-j", "ACCEPT")...)
    }

acceptIPVSTraffic 在NAT表的KUBE-SERVICE链最后添加对所有目地址为ipvs虚拟服务的流量ACCEPT规则(此规则应放置于KUBE-SERVICE的最底部)。默认服务类型clusterip则生成规则-A KUBE-SERVICE -m set --match-set KUBE-CLUSTER-IP dst,dst -j ACCEPT,如果有服务类型为LoadBalancer则生成规则-A KUBE-SERVICE -m set --match-set KUBE-LOAD-BALANCER dst,dst -j ACCEPT

pkg/proxy/ipvs/proxier.go:1397

proxier.acceptIPVSTraffic()
//  -A KUBE-SERVICE -m set --match-set KUBE-CLUSTER-IP dst,dst -j ACCEPT
//  -A KUBE-SERVICE -m set --match-set KUBE-LOAD-BALANCER dst,dst -j ACCEPT
func (proxier *Proxier) acceptIPVSTraffic() {
    sets := []string{kubeClusterIPSet, kubeLoadBalancerSet}
    for _, set := range sets {
        var matchType string
        if !proxier.ipsetList[set].isEmpty() {
            switch proxier.ipsetList[set].SetType {
            case utilipset.BitmapPort:
                matchType = "dst"
            default:
                matchType = "dst,dst"   //目标ip,目标端口
            }
            writeLine(proxier.natRules, []string{
                "-A", string(kubeServicesChain),
                "-m", "set", "--match-set", set, matchType,
                "-j", "ACCEPT",
            }...)
        }
    }
}

增加masqueradeMark,允许NodePort流量转发(即使默认FORWARD规则策略不允许)。

pkg/proxy/ipvs/proxier.go:1361

 // -A KUBE-FORWARD -m comment --comment "kubernetes forwarding rules" -m mark --mark 0x4000 -j ACCEPT
    writeLine(proxier.filterRules,
        "-A", string(KubeForwardChain),
        "-m", "comment", "--comment", `"kubernetes forwarding rules"`,
        "-m", "mark", "--mark", proxier.masqueradeMark,
        "-j", "ACCEPT",
    )

clusterCIDR被指定时生成两条filter表KUBE-FORWARD链规则数据,接受源或目标来自一个pod流量。(注:kube-proxy组件配置--cluster-dir参数指定集群中pod使用的网段)

pkg/proxy/ipvs/proxier.go:1369

    if len(proxier.clusterCIDR) != 0 {
    // 两条规则确保kubernetes forward规则定义的初始包被接受(clusterCIDR所指定的源或目标流量)

    // -A KUBE-FORWARD -s $clusterCIDR -m -comment --comment "kubernetes forwarding conntrack pod source rule" -m conntrack --ctstate RELATED,ESTABLISHED -j ACCEPT
        writeLine(proxier.filterRules,
            "-A", string(KubeForwardChain),
            "-s", proxier.clusterCIDR,
            "-m", "comment", "--comment", `"kubernetes forwarding conntrack pod source rule"`,
            "-m", "conntrack",
            "--ctstate", "RELATED,ESTABLISHED",
            "-j", "ACCEPT",
        )
    // -A KUBE-FORWARD  -m -comment --comment "kubernetes forwarding conntrack pod source rule" -d $clusterCIDR -m conntrack --ctstate RELATED,ESTABLISHED -j ACCEPT
        writeLine(proxier.filterRules,
            "-A", string(KubeForwardChain),
            "-m", "comment", "--comment", `"kubernetes forwarding conntrack pod destination rule"`,
            "-d", proxier.clusterCIDR,
            "-m", "conntrack",
            "--ctstate", "RELATED,ESTABLISHED",
            "-j", "ACCEPT",
        )
    }

4.7. 刷新 iptables 规则

pkg/proxy/ipvs/proxier.go:1186

    // 合并iptables规则
    proxier.iptablesData.Reset()
    proxier.iptablesData.Write(proxier.natChains.Bytes())
    proxier.iptablesData.Write(proxier.natRules.Bytes())
    proxier.iptablesData.Write(proxier.filterChains.Bytes())
    proxier.iptablesData.Write(proxier.filterRules.Bytes())

    klog.V(5).Infof("Restoring iptables rules: %s", proxier.iptablesData.Bytes())
 // 基于iptables格式化规则数据,使用iptables-restore刷新iptables规则
    err = proxier.iptables.RestoreAll(proxier.iptablesData.Bytes(), utiliptables.NoFlushTables, utiliptables.RestoreCounters)
    if err != nil {
        klog.Errorf("Failed to execute iptables-restore: %v\nRules:\n%s", err, proxier.iptablesData.Bytes())
        // Revert new local ports.
        utilproxy.RevertPorts(replacementPortsMap, proxier.portsMap)
        return
    }

ipvs-mode Proxier整个逻辑实现已分析完,其关键逻辑即syncProxyRules(){…}内代码,其中还有一些细节技术未展开叙述,如几个关键的依赖底层技术ipset的实现runner、ipvs路由(VS/RS)操作基于netlink机制通迅机制的实现等,因篇幅过长,后续再看具体情况补充。

~本文 END~

Copyright © farmer.hutao@outlook.com 2019 all right reserved,powered by Gitbook该文件修订时间: 2019-06-20 14:58:28

results matching ""

    No results matching ""

    results matching ""

      No results matching ""