alertmanager源码阅读 - 启动分析

2019独角兽企业重金招聘Python工程师标准>>>

笔者使用的alertmanager版本为0.9.1，

软件地址：https://github.com/prometheus/alertmanager

初始化

关于获取命令行参数，alertmanager和prometheus不同的是，am使用原生的flag库来获取。

configFile      = flag.String("config.file", "alertmanager.yml", "Alertmanager configuration file name.")
dataDir         = flag.String("storage.path", "data/", "Base path for data storage.")

由于目前am压力还小，没有使用HA方案，基于分布式通信协议gossip的mesh及相关的功能暂时不进行深入阅读。

nlog初始化

先看下nlog的定义nflog/nflog.go

type nlog struct {
	logger    log.Logger
	metrics   *metrics
	now       func() time.Time
	retention time.Duration
	runInterval time.Duration
	snapf       string
	stopc       chan struct{}
	done        func()
	mtx sync.RWMutex
	st  gossipData
}

接下来就是nlog的初始化，[]nflog.Option中每一个返回的都是入参为nlog的方法。

notificationLogOpts := []nflog.Option{
		nflog.WithRetention(*retention),
		nflog.WithSnapshot(filepath.Join(*dataDir, "nflog")),
		nflog.WithMaintenance(15*time.Minute, stopc, wg.Done),
		nflog.WithMetrics(prometheus.DefaultRegisterer),
		nflog.WithLogger(log.Base().With("component", "nflog")),
	}
notificationLog, err := nflog.New(notificationLogOpts...)

nflog.New()会执行nflog.Option方法来最终完成nlog初始化，nflog/nflog.go

func New(opts ...Option) (Log, error) {
    l := &nlog{
		logger: log.NewNopLogger(),
		now:    utcNow,
		st:     map[string]*pb.MeshEntry{},
	}
	for _, o := range opts {
		if err := o(l); err != nil {
			return nil, err
		}
	}

    go l.run()
}

marker与silences

marker := types.NewMarker()
silenceOpts := silence.Options{
		SnapshotFile: filepath.Join(*dataDir, "silences"),
		Retention:    *retention,
		Logger:       log.Base().With("component", "silences"),
		Metrics:      prometheus.DefaultRegisterer,
	}
silences, err := silence.New(silenceOpts)

go func() {
		silences.Maintenance(15*time.Minute, filepath.Join(*dataDir, "silences"), stopc)
		wg.Done()
	}()  //每隔15分钟将silences数据写入硬盘
alerts, err := mem.NewAlerts(marker, 30*time.Minute, *dataDir)

func NewAlerts(m types.Marker, intervalGC time.Duration, path string) (*Alerts, error) {
	a := &Alerts{
		alerts:     map[model.Fingerprint]*types.Alert{},
		marker:     m,
		intervalGC: intervalGC,
		stopGC:     make(chan struct{}),
		listeners:  map[int]chan *types.Alert{},
		next:       0,
	}
	go a.runGC() //每隔30分钟将已经解决的alert删除

	return a, nil
}

api

var (
		inhibitor *inhibit.Inhibitor
		tmpl      *template.Template
		pipeline  notify.Stage
		disp      *dispatch.Dispatcher
	)
apiv := api.New(
		alerts,
		silences,
		func(matchers []*labels.Matcher) dispatch.AlertOverview {
			return disp.Groups(matchers)
		},
		marker.Status,
		mrouter,
	)
	apiv.SetMarker(marker)
    apiv.Register(router.WithPrefix("/api")) //在api的uri增加/api前缀

配置

定义了reload方法来进行读取配置和启动相关的方法，并进行reload的首次调用

reload := func() (err error) {
    conf, plainCfg, err := config.LoadFile(*configFile)
    err = apiv.Update(conf, time.Duration(conf.Global.ResolveTimeout))
    inhibitor = inhibit.NewInhibitor(alerts, conf.InhibitRules, marker)
		pipeline = notify.BuildPipeline(
			conf.Receivers,
			tmpl,
			waitFunc,
			inhibitor,
			silences,
			notificationLog,
			marker,
		)
		disp = dispatch.NewDispatcher(alerts, dispatch.NewRoute(conf.Route, nil), pipeline, marker, timeoutFunc)
		go disp.Run()
		go inhibitor.Run()
}

if err := reload(); err != nil {
		os.Exit(1)
	}

更新配置

定义webReload chan，使用其阻塞机制，在调用reload的api时，或者向进程发送SIGHUP信号进行reload()操作

webReload := make(chan struct{})
ui.Register(router, webReload)
var (
		hup      = make(chan os.Signal)
		hupReady = make(chan bool)
		term     = make(chan os.Signal)
	)
	signal.Notify(hup, syscall.SIGHUP)
	signal.Notify(term, os.Interrupt, syscall.SIGTERM)
	go func() {
		<-hupReady
		for {
			select {
			case <-hup:   //接收SIGHUP信号
			case <-webReload: //api接口调用时会向此chan写入数据
			}
			reload()
		}
	}()

api更新的接口代码

ui/web.go

r.Post("/-/reload", func(w http.ResponseWriter, req *http.Request) {
		w.Write([]byte("Reloading configuration file..."))
		reloadCh <- struct{}{}
	})

转载于:https://my.oschina.net/sannychan/blog/1601196