6. informer源码分析-初始化与启动分析
informer架构
先来回忆一下informer的架构。
- Reflector:Reflector 从 kube-apiserver 中 list&watch 资源对象,然后调用 DeltaFIFO 的 Add/Update/Delete/Replace 方法将资源对象及其变化包装成 Delta 并将其丢到 DeltaFIFO 中;
- DeltaFIFO:DeltaFIFO 中存储着一个 map 和一个 queue,即
map[object key]Deltas
以及 object key 的 queue,Deltas 为 Delta 的切片类型,Delta 装有对象及对象的变化类型(Added/Updated/Deleted/Sync) ,Reflector 负责 DeltaFIFO 的输入,Controller 负责处理 DeltaFIFO 的输出; - Controller:Controller 从 DeltaFIFO 的 queue 中 pop 一个 object key 出来,并获取其关联的 Deltas 出来进行处理,遍历 Deltas,根据对象的变化更新 Indexer 中的本地内存缓存,并通知 Processor,相关对象有变化事件发生;
- Processor:Processor 根据对象的变化事件类型,调用相应的 ResourceEventHandler 来处理对象的变化;
- Indexer:Indexer 中有 informer 维护的指定资源对象的相对于 etcd 数据的一份本地内存缓存,可通过该缓存获取资源对象,以减少对 apiserver、对 etcd 的请求压力;
- ResourceEventHandler:用户根据自身处理逻辑需要,注册自定义的的 ResourceEventHandler,当对象发生变化时,将触发调用对应类型的 ResourceEventHandler 来做处理。
<br>
概述
... factory := informers.NewSharedInformerFactory(client, 30*time.Second) podInformer := factory.Core().V1().Pods() informer := podInformer.Informer() ... go factory.Start(stopper) ... if !cache.WaitForCacheSync(stopper, informer.HasSynced) { runtime.HandleError(fmt.Errorf("Timed out waiting for caches to sync")) return } ...
上一节有列举了informer的使用代码,注意看到示例代码中的下面这段代码,做了informer初始化与启动,其中包括:(1)informers.NewSharedInformerFactory:初始化informer factory;(2)podInformer.Informer:初始化pod informer;(3)factory.Start:启动informer factory;(4)cache.WaitForCacheSync:等待list操作获取到的对象都同步到informer本地缓存Indexer中;
下面也将根据这四部分进行informer的初始化与启动分析。
<br>
1.SharedInformerFactory的初始化
1.1 sharedInformerFactory结构体
先来看下 sharedInformerFactory 结构体,看下里面有哪些属性。
看到几个比较重要的属性:
(1)client:连接 k8s 的 clientSet;
(2)informers:是个 map,可以装各个对象的 informer;
(3)startedInformers:记录已经启动的 informer;
// staging/src/k8s.io/client-go/informers/factory.go type sharedInformerFactory struct { client kubernetes.Interface namespace string tweakListOptions internalinterfaces.TweakListOptionsFunc lock sync.Mutex defaultResync time.Duration customResync map[reflect.Type]time.Duration informers map[reflect.Type]cache.SharedIndexInformer // startedInformers is used for tracking which informers have been started. // This allows Start() to be called multiple times safely. startedInformers map[reflect.Type]bool }
1.2 NewSharedInformerFactory
NewSharedInformerFactory 方法用于初始化 informer factory,主要是初始化并返回 sharedInformerFactory 结构体。
// staging/src/k8s.io/client-go/informers/factory.go func NewSharedInformerFactory(client kubernetes.Interface, defaultResync time.Duration) SharedInformerFactory { return NewSharedInformerFactoryWithOptions(client, defaultResync) } func NewFilteredSharedInformerFactory(client kubernetes.Interface, defaultResync time.Duration, namespace string, tweakListOptions internalinterfaces.TweakListOptionsFunc) SharedInformerFactory { return NewSharedInformerFactoryWithOptions(client, defaultResync, WithNamespace(namespace), WithTweakListOptions(tweakListOptions)) } func NewSharedInformerFactoryWithOptions(client kubernetes.Interface, defaultResync time.Duration, options ...SharedInformerOption) SharedInformerFactory { factory := &sharedInformerFactory{ client: client, namespace: v1.NamespaceAll, defaultResync: defaultResync, informers: make(map[reflect.Type]cache.SharedIndexInformer), startedInformers: make(map[reflect.Type]bool), customResync: make(map[reflect.Type]time.Duration), } // Apply all options for _, opt := range options { factory = opt(factory) } return factory }
<br>
2.对象informer的初始化
上一节有列举了 informer 的使用代码,注意看到示例代码中的下面这段代码,这里利用了工厂方法设计模式,podInformer.Informer()
即初始化了 sharedInformerFactory 中的 pod 的 informer,具体调用关系可自行看如下代码,比较简单,这里不再展开分析。
// 初始化informer factory以及pod informer factory := informers.NewSharedInformerFactory(client, 30*time.Second) podInformer := factory.Core().V1().Pods() informer := podInformer.Informer()
2.1 podInformer.Informer
Informer 方法中调用了 f.factory.InformerFor
方法来做 pod informer 的初始化。
// k8s.io/client-go/informers/core/v1/pod.go func (f *podInformer) Informer() cache.SharedIndexInformer { return f.factory.InformerFor(&corev1.Pod{}, f.defaultInformer) }
2.2 f.factory.InformerFor
Informer 方法中调用了 f.factory.InformerFor
方法来做 pod informer 的初始化,并传入 f.defaultInformer
作为 newFunc,而在 f.factory.InformerFor
方法中,调用 newFunc 来初始化 informer。
这里也可以看到,其实 informer 初始化后会存储进 map f.informers[informerType]
中,即存储进 sharedInformerFactory 结构体的 informers 属性中,方便共享使用。
// staging/src/k8s.io/client-go/informers/factory.go func (f *sharedInformerFactory) InformerFor(obj runtime.Object, newFunc internalinterfaces.NewInformerFunc) cache.SharedIndexInformer { f.lock.Lock() defer f.lock.Unlock() informerType := reflect.TypeOf(obj) informer, exists := f.informers[informerType] if exists { return informer } resyncPeriod, exists := f.customResync[informerType] if !exists { resyncPeriod = f.defaultResync } informer = newFunc(f.client, resyncPeriod) f.informers[informerType] = informer return informer }
2.3 newFunc/f.defaultInformer
defaultInformer 方法中,调用了 NewFilteredPodInformer 方法来初始化 pod informer,最终初始化并返回 sharedIndexInformer 结构体。
// k8s.io/client-go/informers/core/v1/pod.go func (f *podInformer) defaultInformer(client kubernetes.Interface, resyncPeriod time.Duration) cache.SharedIndexInformer { return NewFilteredPodInformer(client, f.namespace, resyncPeriod, cache.Indexers{cache.NamespaceIndex: cache.MetaNamespaceIndexFunc}, f.tweakListOptions) } func NewFilteredPodInformer(client kubernetes.Interface, namespace string, resyncPeriod time.Duration, indexers cache.Indexers, tweakListOptions internalinterfaces.TweakListOptionsFunc) cache.SharedIndexInformer { return cache.NewSharedIndexInformer( &cache.ListWatch{ ListFunc: func(options metav1.ListOptions) (runtime.Object, error) { if tweakListOptions != nil { tweakListOptions(&options) } return client.CoreV1().Pods(namespace).List(options) }, WatchFunc: func(options metav1.ListOptions) (watch.Interface, error) { if tweakListOptions != nil { tweakListOptions(&options) } return client.CoreV1().Pods(namespace).Watch(options) }, }, &corev1.Pod{}, resyncPeriod, indexers, ) } func NewSharedIndexInformer(lw ListerWatcher, objType runtime.Object, defaultEventHandlerResyncPeriod time.Duration, indexers Indexers) SharedIndexInformer { realClock := &clock.RealClock{} sharedIndexInformer := &sharedIndexInformer{ processor: &sharedProcessor{clock: realClock}, indexer: NewIndexer(DeletionHandlingMetaNamespaceKeyFunc, indexers), listerWatcher: lw, objectType: objType, resyncCheckPeriod: defaultEventHandlerResyncPeriod, defaultEventHandlerResyncPeriod: defaultEventHandlerResyncPeriod, cacheMutationDetector: NewCacheMutationDetector(fmt.Sprintf("%T", objType)), clock: realClock, } return sharedIndexInformer }
2.4 sharedIndexInformer结构体
sharedIndexInformer 结构体中重点看到以下几个属性:
(1)indexer:对应着 informer 中的部件 Indexer,Indexer 中有 informer 维护的指定资源对象的相对于 etcd 数据的一份本地内存缓存,可通过该缓存获取资源对象,以减少对 apiserver、对 etcd 的请求压力;
(2)controller:对应着 informer 中的部件 Controller,Controller 从 DeltaFIFO 中 pop Deltas 出来处理,根据对象的变化更新 Indexer 中的本地内存缓存,并通知 Processor,相关对象有变化事件发生;
(3)processor:对应着 informer 中的部件 Processor,Processor 根据对象的变化事件类型,调用相应的 ResourceEventHandler 来处理对象的变化;
// staging/src/k8s.io/client-go/tools/cache/shared_informer.go type sharedIndexInformer struct { indexer Indexer controller Controller processor *sharedProcessor cacheMutationDetector CacheMutationDetector // This block is tracked to handle late initialization of the controller listerWatcher ListerWatcher objectType runtime.Object // resyncCheckPeriod is how often we want the reflector's resync timer to fire so it can call // shouldResync to check if any of our listeners need a resync. resyncCheckPeriod time.Duration // defaultEventHandlerResyncPeriod is the default resync period for any handlers added via // AddEventHandler (i.e. they don't specify one and just want to use the shared informer's default // value). defaultEventHandlerResyncPeriod time.Duration // clock allows for testability clock clock.Clock started, stopped bool startedLock sync.Mutex // blockDeltas gives a way to stop all event distribution so that a late event handler // can safely join the shared informer. blockDeltas sync.Mutex }
Indexer接口与cache结构体cache结构体为Indexer接口的实现;
// staging/src/k8s.io/client-go/tools/cache/store.go type cache struct { cacheStorage ThreadSafeStore keyFunc KeyFunc }
threadSafeMap struct 是 ThreadSafeStore 接口的一个实现,其最重要的一个属性便是 items 了,items 是用 map 构建的键值对,资源对象都存在 items 这个 map 中,key 根据资源对象来算出,value 为资源对象本身,这里的 items 即为 informer 的本地缓存了,而 indexers 与 indices 属性则与索引功能有关。
// staging/src/k8s.io/client-go/tools/cache/thread_safe_store.go type threadSafeMap struct { lock sync.RWMutex items map[string]interface{} // indexers maps a name to an IndexFunc indexers Indexers // indices maps a name to an Index indices Indices }
关于 Indexer 的详细分析会在后续有专门的文章做分析,这里不展开分析;
controller 结构体
而 controller 结构体则包含了 informer 中的主要部件 Reflector 以及 DeltaFIFO;(1)Reflector:Reflector 从 kube-apiserver 中 list&watch 资源对象,然后将对象的变化包装成 Delta 并将其丢到 DeltaFIFO 中;(2)DeltaFIFO:DeltaFIFO 存储着 map[object key]Deltas
以及 object key 的 queue,Delta 装有对象及对象的变化类型 ,Reflector 负责 DeltaFIFO 的输入,Controller 负责处理 DeltaFIFO 的输出;
// staging/src/k8s.io/client-go/tools/cache/controller.go type controller struct { config Config reflector *Reflector reflectorMutex sync.RWMutex clock clock.Clock } type Config struct { // The queue for your objects; either a FIFO or // a DeltaFIFO. Your Process() function should accept // the output of this Queue's Pop() method. Queue ... }
<br>
3.启动sharedInformerFactory
sharedInformerFactory.Start
为 informer factory 的启动方法,其主要逻辑为循环遍历 informers,然后跑 goroutine 调用 informer.Run
来启动 sharedInformerFactory 中存储的各个 informer。
// staging/src/k8s.io/client-go/informers/factory.go func (f *sharedInformerFactory) Start(stopCh <-chan struct{}) { f.lock.Lock() defer f.lock.Unlock() for informerType, informer := range f.informers { if !f.startedInformers[informerType] { go informer.Run(stopCh) f.startedInformers[informerType] = true } } }
sharedIndexInformer.Run
sharedIndexInformer.Run
用于启动 informer,主要逻辑为:
(1)调用 NewDeltaFIFO,初始化 DeltaFIFO;
(2)构建 Config 结构体,这里留意下 Process 属性,赋值了 s.HandleDeltas
,后面会分析到该方法;
(3)调用 New,利用 Config 结构体来初始化 controller;
(4)调用 s.processor.run
,启动 processor;
(5)调用 s.controller.Run
,启动 controller;
// staging/src/k8s.io/client-go/tools/cache/shared_informer.go func (s *sharedIndexInformer) Run(stopCh <-chan struct{}) { defer utilruntime.HandleCrash() // 初始化DeltaFIFO fifo := NewDeltaFIFO(MetaNamespaceKeyFunc, s.indexer) // 构建Config结构体 cfg := &Config{ Queue: fifo, ListerWatcher: s.listerWatcher, ObjectType: s.objectType, FullResyncPeriod: s.resyncCheckPeriod, RetryOnError: false, ShouldResync: s.processor.shouldResync, Process: s.HandleDeltas, } func() { s.startedLock.Lock() defer s.startedLock.Unlock() // 初始化controller s.controller = New(cfg) s.controller.(*controller).clock = s.clock s.started = true }() // Separate stop channel because Processor should be stopped strictly after controller processorStopCh := make(chan struct{}) var wg wait.Group defer wg.Wait() // Wait for Processor to stop defer close(processorStopCh) // Tell Processor to stop wg.StartWithChannel(processorStopCh, s.cacheMutationDetector.Run) // 启动processor wg.StartWithChannel(processorStopCh, s.processor.run) defer func() { s.startedLock.Lock() defer s.startedLock.Unlock() s.stopped = true // Don't want any new listeners }() // 启动controller s.controller.Run(stopCh) }
3.1 New
New 函数初始化了 controller 并 return。
// staging/src/k8s.io/client-go/tools/cache/controller.go func New(c *Config) Controller { ctlr := &controller{ config: *c, clock: &clock.RealClock{}, } return ctlr }
3.2 s.processor.run
s.processor.run
启动了 processor,其中注意到 listener.run
与 listener.pop
两个核心方法即可,暂时没有用到,等下面用到他们的时候再做分析。
// staging/src/k8s.io/client-go/tools/cache/shared_informer.go func (p *sharedProcessor) run(stopCh <-chan struct{}) { func() { p.listenersLock.RLock() defer p.listenersLock.RUnlock() for _, listener := range p.listeners { p.wg.Start(listener.run) p.wg.Start(listener.pop) } p.listenersStarted = true }() <-stopCh p.listenersLock.RLock() defer p.listenersLock.RUnlock() for _, listener := range p.listeners { close(listener.addCh) // Tell .pop() to stop. .pop() will tell .run() to stop } p.wg.Wait() // Wait for all .pop() and .run() to stop }
3.3 controller.Run
controller.Run
为 controller 的启动方法,这里主要看到几个点:
(1)调用 NewReflector,初始化 Reflector;
(2)调用 r.Run
,实际上是调用了 Reflector 的启动方法来启动Reflector;
(3)调用 c.processLoop
,开始 controller 的核心处理;
// k8s.io/client-go/tools/cache/controller.go func (c *controller) Run(stopCh <-chan struct{}) { defer utilruntime.HandleCrash() go func() { <-stopCh c.config.Queue.Close() }() r := NewReflector( c.config.ListerWatcher, c.config.ObjectType, c.config.Queue, c.config.FullResyncPeriod, ) r.ShouldResync = c.config.ShouldResync r.clock = c.clock c.reflectorMutex.Lock() c.reflector = r c.reflectorMutex.Unlock() var wg wait.Group defer wg.Wait() wg.StartWithChannel(stopCh, r.Run) wait.Until(c.processLoop, time.Second, stopCh) }
3.3.1 Reflector结构体
先来看到 Reflector 结构体,这里重点看到以下属性:
(1)expectedType:放到 Store 中(即 DeltaFIFO 中)的对象类型;
(2)store:store 会赋值为 DeltaFIFO,具体可以看之前的 informer 初始化与启动分析即可得知,这里不再展开分析;
(3)listerWatcher:存放 list 方法和 watch 方法的 ListerWatcher interface 实现;
// k8s.io/client-go/tools/cache/reflector.go type Reflector struct { ... expectedType reflect.Type store Store listerWatcher ListerWatcher ... }
3.3.2 r.Run/Reflector.Run
Reflector.Run
方法中启动了 Reflector,而 Reflector 的核心处理逻辑为从 kube-apiserver 处做 list&watch 操作,然后将得到的对象封装存储进 DeltaFIFO 中。
// staging/src/k8s.io/client-go/tools/cache/reflector.go func (r *Reflector) Run(stopCh <-chan struct{}) { klog.V(3).Infof("Starting reflector %v (%s) from %s", r.expectedTypeName, r.resyncPeriod, r.name) wait.Until(func() { if err := r.ListAndWatch(stopCh); err != nil { utilruntime.HandleError(err) } }, r.period, stopCh) }
3.3.3 controller.processLoop
controller 的核心处理方法 processLoop 中,最重要的逻辑是循环调用 c.config.Queue.Pop
将 DeltaFIFO 中的队头元素给 pop 出来,然后调用 c.config.Process
方法来做处理,当处理出错时,再调用 c.config.Queue.AddIfNotPresent
将对象重新加入到 DeltaFIFO 中去。
// k8s.io/client-go/tools/cache/controller.go func (c *controller) processLoop() { for { obj, err := c.config.Queue.Pop(PopProcessFunc(c.config.Process)) if err != nil { if err == ErrFIFOClosed { return } if c.config.RetryOnError { // This is the safe way to re-enqueue. c.config.Queue.AddIfNotPresent(obj) } } } }
3.3.4 c.config.Process/sharedIndexInformer.HandleDeltas
根据前面 sharedIndexInformer.Run
方法的分析中可以得知, c.config.Process
其实就是 sharedIndexInformer.HandleDeltas
。
HandleDeltas 方法中,将从 DeltaFIFO 中 pop 出来的对象以及类型,相应的在 indexer 中做添加、更新、删除操作,并调用 s.processor.distribute
通知自定义的 ResourceEventHandler。
// staging/src/k8s.io/client-go/tools/cache/shared_informer.go func (s *sharedIndexInformer) HandleDeltas(obj interface{}) error { s.blockDeltas.Lock() defer s.blockDeltas.Unlock() // from oldest to newest for _, d := range obj.(Deltas) { switch d.Type { case Sync, Added, Updated: isSync := d.Type == Sync s.cacheMutationDetector.AddObject(d.Object) if old, exists, err := s.indexer.Get(d.Object); err == nil && exists { if err := s.indexer.Update(d.Object); err != nil { return err } s.processor.distribute(updateNotification{oldObj: old, newObj: d.Object}, isSync) } else { if err := s.indexer.Add(d.Object); err != nil { return err } s.processor.distribute(addNotification{newObj: d.Object}, isSync) } case Deleted: if err := s.indexer.Delete(d.Object); err != nil { return err } s.processor.distribute(deleteNotification{oldObj: d.Object}, false) } } return nil }
怎么通知到自定义的ResourceEventHandler呢?继续往下看。
3.3.5 sharedIndexInformer.processor.distribute
可以看到 distribute 方法最终是将构造好的 addNotification、updateNotification、deleteNotification 对象写入到 p.addCh
中。
// staging/src/k8s.io/client-go/tools/cache/shared_informer.go func (p *sharedProcessor) distribute(obj interface{}, sync bool) { p.listenersLock.RLock() defer p.listenersLock.RUnlock() if sync { for _, listener := range p.syncingListeners { listener.add(obj) } } else { for _, listener := range p.listeners { listener.add(obj) } } } func (p *processorListener) add(notification interface{}) { p.addCh <- notification }
到这里,processor 中的 listener.pop
以及 listener.run
方法终于派上了用场,继续往下看。
3.3.6 listener.pop
分析 processorListener 的 pop 方法可以得知,其逻辑实际上就是将 p.addCh 中的对象给拿出来,然后丢进了 p.nextCh
中。那么谁来处理 p.nextCh
呢?继续往下看。
// staging/src/k8s.io/client-go/tools/cache/shared_informer.go func (p *processorListener) pop() { defer utilruntime.HandleCrash() defer close(p.nextCh) // Tell .run() to stop var nextCh chan<- interface{} var notification interface{} for { select { case nextCh <- notification: // Notification dispatched var ok bool notification, ok = p.pendingNotifications.ReadOne() if !ok { // Nothing to pop nextCh = nil // Disable this select case } case notificationToAdd, ok := <-p.addCh: if !ok { return } if notification == nil { // No notification to pop (and pendingNotifications is empty) // Optimize the case - skip adding to pendingNotifications notification = notificationToAdd nextCh = p.nextCh } else { // There is already a notification waiting to be dispatched p.pendingNotifications.WriteOne(notificationToAdd) } } } }
3.3.7 listener.run
在 processorListener 的 run 方法中,将循环读取 p.nextCh
,判断对象类型,是 updateNotification 则调用 p.handler.OnUpdate
方法,是 addNotification 则调用 p.handler.OnAdd
方法,是 deleteNotification 则调用 p.handler.OnDelete
方法做处理。
// staging/src/k8s.io/client-go/tools/cache/shared_informer.go func (p *processorListener) run() { // this call blocks until the channel is closed. When a panic happens during the notification // we will catch it, **the offending item will be skipped!**, and after a short delay (one second) // the next notification will be attempted. This is usually better than the alternative of never // delivering again. stopCh := make(chan struct{}) wait.Until(func() { // this gives us a few quick retries before a long pause and then a few more quick retries err := wait.ExponentialBackoff(retry.DefaultRetry, func() (bool, error) { for next := range p.nextCh { switch notification := next.(type) { case updateNotification: p.handler.OnUpdate(notification.oldObj, notification.newObj) case addNotification: p.handler.OnAdd(notification.newObj) case deleteNotification: p.handler.OnDelete(notification.oldObj) default: utilruntime.HandleError(fmt.Errorf("unrecognized notification: %T", next)) } } // the only way to get here is if the p.nextCh is empty and closed return true, nil }) // the only way to get here is if the p.nextCh is empty and closed if err == nil { close(stopCh) } }, 1*time.Minute, stopCh) }
而 p.handler.OnUpdate、p.handler.OnAdd、p.handler.OnDelete
方法实际上就是自定义的的 ResourceEventHandlerFuncs 了。
informer.AddEventHandler(cache.ResourceEventHandlerFuncs{ AddFunc: onAdd, UpdateFunc: onUpdate, DeleteFunc: onDelete, }) // staging/src/k8s.io/client-go/tools/cache/controller.go type ResourceEventHandlerFuncs struct { AddFunc func(obj interface{}) UpdateFunc func(oldObj, newObj interface{}) DeleteFunc func(obj interface{}) } func (r ResourceEventHandlerFuncs) OnAdd(obj interface{}) { if r.AddFunc != nil { r.AddFunc(obj) } } func (r ResourceEventHandlerFuncs) OnUpdate(oldObj, newObj interface{}) { if r.UpdateFunc != nil { r.UpdateFunc(oldObj, newObj) } } func (r ResourceEventHandlerFuncs) OnDelete(obj interface{}) { if r.DeleteFunc != nil { r.DeleteFunc(obj) } }
<br>
4.cache.WaitForCacheSync(stopper, informer.HasSynced)
可以看出在 cache.WaitForCacheSync
方法中,实际上是调用方法入参 cacheSyncs ...InformerSynced
来判断 cache 是否同步完成(即调用 informer.HasSynced
方法),而这里说的 cache 同步完成,意思是等待 informer 从 kube-apiserver 同步资源完成,即 informer 的 list 操作获取的对象都存入到 informer 中的 indexer 本地缓存中;
// staging/src/k8s.io/client-go/tools/cache/shared_informer.go func WaitForCacheSync(stopCh <-chan struct{}, cacheSyncs ...InformerSynced) bool { err := wait.PollImmediateUntil(syncedPollPeriod, func() (bool, error) { for _, syncFunc := range cacheSyncs { if !syncFunc() { return false, nil } } return true, nil }, stopCh) if err != nil { klog.V(2).Infof("stop requested") return false } klog.V(4).Infof("caches populated") return true }
4.1 informer.HasSynced
HasSynced 方法实际上是调用了 sharedIndexInformer.controller.HasSynced
方法;
// staging/src/k8s.io/client-go/tools/cache/shared_informer.go func (s *sharedIndexInformer) HasSynced() bool { s.startedLock.Lock() defer s.startedLock.Unlock() if s.controller == nil { return false } return s.controller.HasSynced() }
s.controller.HasSynced
这里的 c.config.Queue.HasSynced()
方法,实际上是指 DeltaFIFO 的 HasSynced 方法,会在 DeltaFIFO 的分析中再详细分析,这里只需要知道当 informer 的 list 操作获取的对象都存入到 informer 中的 indexer 本地缓存中则返回 true 即可;
// staging/src/k8s.io/client-go/tools/cache/controller.go func (c *controller) HasSynced() bool { return c.config.Queue.HasSynced() }
4.2 sharedInformerFactory.WaitForCacheSync
可以顺带看下 sharedInformerFactory.WaitForCacheSync
方法,其实际上是遍历 factory 中的所有 informer,调用 cache.WaitForCacheSync
,然后传入每个 informer 的 HasSynced 方法作为入参;
// staging/src/k8s.io/client-go/informers/factory.go func (f *sharedInformerFactory) WaitForCacheSync(stopCh <-chan struct{}) map[reflect.Type]bool { informers := func() map[reflect.Type]cache.SharedIndexInformer { f.lock.Lock() defer f.lock.Unlock() informers := map[reflect.Type]cache.SharedIndexInformer{} for informerType, informer := range f.informers { if f.startedInformers[informerType] { informers[informerType] = informer } } return informers }() res := map[reflect.Type]bool{} for informType, informer := range informers { res[informType] = cache.WaitForCacheSync(stopCh, informer.HasSynced) } return res }
至此,整个 informer 的初始化与启动的分析就结束了,后面会对 informer 中的各个核心部件进行详细分析,敬请期待。
总结下面用两张图片总结一下 informer 的初始化与启动;
informer初始化
informer启动
Kubernetes源码阅读