PNG  IHDR pHYs   OiCCPPhotoshop ICC profilexڝSgTS=BKKoR RB&*! J!QEEȠQ, !{kּ> H3Q5 B.@ $pd!s#~<<+"x M0B\t8K@zB@F&S`cbP-`'{[! eDh;VEX0fK9-0IWfH  0Q){`##xFW<+*x<$9E[-qWW.(I+6aa@.y24x6_-"bbϫp@t~,/;m%h^ uf@Wp~<5j>{-]cK'Xto(hw?G%fIq^D$.Tʳ?D*A, `6B$BB dr`)B(Ͱ*`/@4Qhp.U=pa( Aa!ڈbX#!H$ ɈQ"K5H1RT UH=r9\F;2G1Q= C7F dt1r=6Ыhڏ>C03l0.B8, c˱" VcϱwE 6wB aAHXLXNH $4 7 Q'"K&b21XH,#/{C7$C2'ITFnR#,4H#dk9, +ȅ3![ b@qS(RjJ4e2AURݨT5ZBRQ4u9̓IKhhitݕNWGw Ljg(gwLӋT071oUX**| J&*/Tު UUT^S}FU3S ԖUPSSg;goT?~YYLOCQ_ cx,!k u5&|v*=9C3J3WRf?qtN (~))4L1e\kXHQG6EYAJ'\'GgSSݧ M=:.kDwn^Loy}/TmG X $ <5qo</QC]@Caaᄑ.ȽJtq]zۯ6iܟ4)Y3sCQ? 0k߬~OCOg#/c/Wװwa>>r><72Y_7ȷOo_C#dz%gA[z|!?:eAAA!h쐭!ΑiP~aa~ 'W?pX15wCsDDDޛg1O9-J5*>.j<74?.fYXXIlK9.*6nl {/]py.,:@LN8A*%w% yg"/6шC\*NH*Mz쑼5y$3,幄'L Lݛ:v m2=:1qB!Mggfvˬen/kY- BTZ(*geWf͉9+̳ې7ᒶKW-X潬j9(xoʿܔĹdff-[n ڴ VE/(ۻCɾUUMfeI?m]Nmq#׹=TR+Gw- 6 U#pDy  :v{vg/jBFS[b[O>zG499?rCd&ˮ/~јѡ򗓿m|x31^VwwO| (hSЧc3- cHRMz%u0`:o_F@8N ' p @8N@8}' p '#@8N@8N pQ9p!i~}|6-ӪG` VP.@*j>[ K^<֐Z]@8N'KQ<Q(`s" 'hgpKB`R@Dqj '  'P$a ( `D$Na L?u80e J,K˷NI'0eݷ(NI'؀ 2ipIIKp`:O'`ʤxB8Ѥx Ѥx $ $P6 :vRNb 'p,>NB 'P]-->P T+*^h& p '‰a ‰ (ĵt#u33;Nt̵'ޯ; [3W ~]0KH1q@8]O2]3*̧7# *p>us p _6]/}-4|t'|Smx= DoʾM×M_8!)6lq':l7!|4} '\ne t!=hnLn (~Dn\+‰_4k)0e@OhZ`F `.m1} 'vp{F`ON7Srx 'D˸nV`><;yMx!IS钦OM)Ե٥x 'DSD6bS8!" ODz#R >S8!7ّxEh0m$MIPHi$IvS8IN$I p$O8I,sk&I)$IN$Hi$I^Ah.p$MIN$IR8I·N "IF9Ah0m$MIN$IR8IN$I 3jIU;kO$ɳN$+ q.x* tEXtComment

Viewing File: /opt/go/pkg/mod/github.com/prometheus/alertmanager@v0.28.0/cluster/delegate.go

// Copyright 2018 Prometheus Team
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

package cluster

import (
	"log/slog"
	"time"

	"github.com/gogo/protobuf/proto"
	"github.com/hashicorp/memberlist"
	"github.com/prometheus/client_golang/prometheus"

	"github.com/prometheus/alertmanager/cluster/clusterpb"
)

const (
	// Maximum number of messages to be held in the queue.
	maxQueueSize = 4096
	fullState    = "full_state"
	update       = "update"
)

// delegate implements memberlist.Delegate and memberlist.EventDelegate
// and broadcasts its peer's state in the cluster.
type delegate struct {
	*Peer

	logger *slog.Logger
	bcast  *memberlist.TransmitLimitedQueue

	messagesReceived     *prometheus.CounterVec
	messagesReceivedSize *prometheus.CounterVec
	messagesSent         *prometheus.CounterVec
	messagesSentSize     *prometheus.CounterVec
	messagesPruned       prometheus.Counter
	nodeAlive            *prometheus.CounterVec
	nodePingDuration     *prometheus.HistogramVec
}

func newDelegate(l *slog.Logger, reg prometheus.Registerer, p *Peer, retransmit int) *delegate {
	bcast := &memberlist.TransmitLimitedQueue{
		NumNodes:       p.ClusterSize,
		RetransmitMult: retransmit,
	}
	messagesReceived := prometheus.NewCounterVec(prometheus.CounterOpts{
		Name: "alertmanager_cluster_messages_received_total",
		Help: "Total number of cluster messages received.",
	}, []string{"msg_type"})
	messagesReceivedSize := prometheus.NewCounterVec(prometheus.CounterOpts{
		Name: "alertmanager_cluster_messages_received_size_total",
		Help: "Total size of cluster messages received.",
	}, []string{"msg_type"})
	messagesSent := prometheus.NewCounterVec(prometheus.CounterOpts{
		Name: "alertmanager_cluster_messages_sent_total",
		Help: "Total number of cluster messages sent.",
	}, []string{"msg_type"})
	messagesSentSize := prometheus.NewCounterVec(prometheus.CounterOpts{
		Name: "alertmanager_cluster_messages_sent_size_total",
		Help: "Total size of cluster messages sent.",
	}, []string{"msg_type"})
	messagesPruned := prometheus.NewCounter(prometheus.CounterOpts{
		Name: "alertmanager_cluster_messages_pruned_total",
		Help: "Total number of cluster messages pruned.",
	})
	gossipClusterMembers := prometheus.NewGaugeFunc(prometheus.GaugeOpts{
		Name: "alertmanager_cluster_members",
		Help: "Number indicating current number of members in cluster.",
	}, func() float64 {
		return float64(p.ClusterSize())
	})
	peerPosition := prometheus.NewGaugeFunc(prometheus.GaugeOpts{
		Name: "alertmanager_peer_position",
		Help: "Position the Alertmanager instance believes it's in. The position determines a peer's behavior in the cluster.",
	}, func() float64 {
		return float64(p.Position())
	})
	healthScore := prometheus.NewGaugeFunc(prometheus.GaugeOpts{
		Name: "alertmanager_cluster_health_score",
		Help: "Health score of the cluster. Lower values are better and zero means 'totally healthy'.",
	}, func() float64 {
		return float64(p.mlist.GetHealthScore())
	})
	messagesQueued := prometheus.NewGaugeFunc(prometheus.GaugeOpts{
		Name: "alertmanager_cluster_messages_queued",
		Help: "Number of cluster messages which are queued.",
	}, func() float64 {
		return float64(bcast.NumQueued())
	})
	nodeAlive := prometheus.NewCounterVec(prometheus.CounterOpts{
		Name: "alertmanager_cluster_alive_messages_total",
		Help: "Total number of received alive messages.",
	}, []string{"peer"},
	)
	nodePingDuration := prometheus.NewHistogramVec(prometheus.HistogramOpts{
		Name:                            "alertmanager_cluster_pings_seconds",
		Help:                            "Histogram of latencies for ping messages.",
		Buckets:                         []float64{.005, .01, .025, .05, .1, .25, .5},
		NativeHistogramBucketFactor:     1.1,
		NativeHistogramMaxBucketNumber:  100,
		NativeHistogramMinResetDuration: 1 * time.Hour,
	}, []string{"peer"},
	)

	messagesReceived.WithLabelValues(fullState)
	messagesReceivedSize.WithLabelValues(fullState)
	messagesReceived.WithLabelValues(update)
	messagesReceivedSize.WithLabelValues(update)
	messagesSent.WithLabelValues(fullState)
	messagesSentSize.WithLabelValues(fullState)
	messagesSent.WithLabelValues(update)
	messagesSentSize.WithLabelValues(update)

	reg.MustRegister(messagesReceived, messagesReceivedSize, messagesSent, messagesSentSize,
		gossipClusterMembers, peerPosition, healthScore, messagesQueued, messagesPruned,
		nodeAlive, nodePingDuration,
	)

	d := &delegate{
		logger:               l,
		Peer:                 p,
		bcast:                bcast,
		messagesReceived:     messagesReceived,
		messagesReceivedSize: messagesReceivedSize,
		messagesSent:         messagesSent,
		messagesSentSize:     messagesSentSize,
		messagesPruned:       messagesPruned,
		nodeAlive:            nodeAlive,
		nodePingDuration:     nodePingDuration,
	}

	go d.handleQueueDepth()

	return d
}

// NodeMeta retrieves meta-data about the current node when broadcasting an alive message.
func (d *delegate) NodeMeta(limit int) []byte {
	return []byte{}
}

// NotifyMsg is the callback invoked when a user-level gossip message is received.
func (d *delegate) NotifyMsg(b []byte) {
	d.messagesReceived.WithLabelValues(update).Inc()
	d.messagesReceivedSize.WithLabelValues(update).Add(float64(len(b)))

	var p clusterpb.Part
	if err := proto.Unmarshal(b, &p); err != nil {
		d.logger.Warn("decode broadcast", "err", err)
		return
	}

	d.mtx.RLock()
	s, ok := d.states[p.Key]
	d.mtx.RUnlock()

	if !ok {
		return
	}
	if err := s.Merge(p.Data); err != nil {
		d.logger.Warn("merge broadcast", "err", err, "key", p.Key)
		return
	}
}

// GetBroadcasts is called when user data messages can be broadcasted.
func (d *delegate) GetBroadcasts(overhead, limit int) [][]byte {
	msgs := d.bcast.GetBroadcasts(overhead, limit)
	d.messagesSent.WithLabelValues(update).Add(float64(len(msgs)))
	for _, m := range msgs {
		d.messagesSentSize.WithLabelValues(update).Add(float64(len(m)))
	}
	return msgs
}

// LocalState is called when gossip fetches local state.
func (d *delegate) LocalState(_ bool) []byte {
	d.mtx.RLock()
	defer d.mtx.RUnlock()
	all := &clusterpb.FullState{
		Parts: make([]clusterpb.Part, 0, len(d.states)),
	}

	for key, s := range d.states {
		b, err := s.MarshalBinary()
		if err != nil {
			d.logger.Warn("encode local state", "err", err, "key", key)
			return nil
		}
		all.Parts = append(all.Parts, clusterpb.Part{Key: key, Data: b})
	}
	b, err := proto.Marshal(all)
	if err != nil {
		d.logger.Warn("encode local state", "err", err)
		return nil
	}
	d.messagesSent.WithLabelValues(fullState).Inc()
	d.messagesSentSize.WithLabelValues(fullState).Add(float64(len(b)))
	return b
}

func (d *delegate) MergeRemoteState(buf []byte, _ bool) {
	d.messagesReceived.WithLabelValues(fullState).Inc()
	d.messagesReceivedSize.WithLabelValues(fullState).Add(float64(len(buf)))

	var fs clusterpb.FullState
	if err := proto.Unmarshal(buf, &fs); err != nil {
		d.logger.Warn("merge remote state", "err", err)
		return
	}
	d.mtx.RLock()
	defer d.mtx.RUnlock()
	for _, p := range fs.Parts {
		s, ok := d.states[p.Key]
		if !ok {
			d.logger.Warn("unknown state key", "len", len(buf), "key", p.Key)
			continue
		}
		if err := s.Merge(p.Data); err != nil {
			d.logger.Warn("merge remote state", "err", err, "key", p.Key)
			return
		}
	}
}

// NotifyJoin is called if a peer joins the cluster.
func (d *delegate) NotifyJoin(n *memberlist.Node) {
	d.logger.Debug("NotifyJoin", "node", n.Name, "addr", n.Address())
	d.Peer.peerJoin(n)
}

// NotifyLeave is called if a peer leaves the cluster.
func (d *delegate) NotifyLeave(n *memberlist.Node) {
	d.logger.Debug("NotifyLeave", "node", n.Name, "addr", n.Address())
	d.Peer.peerLeave(n)
}

// NotifyUpdate is called if a cluster peer gets updated.
func (d *delegate) NotifyUpdate(n *memberlist.Node) {
	d.logger.Debug("NotifyUpdate", "node", n.Name, "addr", n.Address())
	d.Peer.peerUpdate(n)
}

// NotifyAlive implements the memberlist.AliveDelegate interface.
func (d *delegate) NotifyAlive(peer *memberlist.Node) error {
	d.nodeAlive.WithLabelValues(peer.Name).Inc()
	return nil
}

// AckPayload implements the memberlist.PingDelegate interface.
func (d *delegate) AckPayload() []byte {
	return []byte{}
}

// NotifyPingComplete implements the memberlist.PingDelegate interface.
func (d *delegate) NotifyPingComplete(peer *memberlist.Node, rtt time.Duration, payload []byte) {
	d.nodePingDuration.WithLabelValues(peer.Name).Observe(rtt.Seconds())
}

// handleQueueDepth ensures that the queue doesn't grow unbounded by pruning
// older messages at regular interval.
func (d *delegate) handleQueueDepth() {
	for {
		select {
		case <-d.stopc:
			return
		case <-time.After(15 * time.Minute):
			n := d.bcast.NumQueued()
			if n > maxQueueSize {
				d.logger.Warn("dropping messages because too many are queued", "current", n, "limit", maxQueueSize)
				d.bcast.Prune(maxQueueSize)
				d.messagesPruned.Add(float64(n - maxQueueSize))
			}
		}
	}
}
Back to Directory=ceiIENDB`