forked from mdehoog/optimism
-
Notifications
You must be signed in to change notification settings - Fork 0
/
metrics.go
118 lines (103 loc) · 3.19 KB
/
metrics.go
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
package op_heartbeat
import (
"fmt"
"strconv"
"sync/atomic"
"time"
lru "github.com/hashicorp/golang-lru/v2"
"github.com/prometheus/client_golang/prometheus"
"github.com/prometheus/client_golang/prometheus/promauto"
"github.com/ethereum-optimism/optimism/op-node/heartbeat"
)
const (
MetricsNamespace = "op_heartbeat"
MinHeartbeatInterval = 10*time.Minute - 10*time.Second
UsersCacheSize = 10_000
)
type Metrics interface {
RecordHeartbeat(payload heartbeat.Payload, ip string)
RecordVersion(version string)
}
type metrics struct {
heartbeats *prometheus.CounterVec
version *prometheus.GaugeVec
sameIP *prometheus.HistogramVec
// Groups heartbeats per unique IP, version and chain ID combination.
// string(IP ++ version ++ chainID) -> *heartbeatEntry
heartbeatUsers *lru.Cache[string, *heartbeatEntry]
}
type heartbeatEntry struct {
// Count number of heartbeats per interval, atomically updated
Count uint64
// Changes once per heartbeat interval
Time time.Time
}
func NewMetrics(r *prometheus.Registry) Metrics {
lruCache, _ := lru.New[string, *heartbeatEntry](UsersCacheSize)
m := &metrics{
heartbeats: promauto.With(r).NewCounterVec(prometheus.CounterOpts{
Namespace: MetricsNamespace,
Name: "heartbeats",
Help: "Counts number of heartbeats by chain ID, version and filtered to unique IPs",
}, []string{
"chain_id",
"version",
}),
version: promauto.With(r).NewGaugeVec(prometheus.GaugeOpts{
Namespace: MetricsNamespace,
Name: "version",
Help: "version pseudo-metrics",
}, []string{
"version",
}),
sameIP: promauto.With(r).NewHistogramVec(prometheus.HistogramOpts{
Namespace: MetricsNamespace,
Name: "heartbeat_same_ip",
Buckets: []float64{1, 2, 4, 8, 16, 32, 64, 128},
Help: "Histogram of events within same heartbeat interval per unique IP, by chain ID and version",
}, []string{
"chain_id",
"version",
}),
heartbeatUsers: lruCache,
}
return m
}
func (m *metrics) RecordHeartbeat(payload heartbeat.Payload, ip string) {
var chainID string
if AllowedChainIDs[payload.ChainID] {
chainID = strconv.FormatUint(payload.ChainID, 10)
} else {
chainID = "unknown"
}
var version string
if AllowedVersions[payload.Version] {
version = payload.Version
} else {
version = "unknown"
}
key := fmt.Sprintf("%s;%s;%s", ip, version, chainID)
now := time.Now()
entry, ok, _ := m.heartbeatUsers.PeekOrAdd(key, &heartbeatEntry{Time: now, Count: 1})
if !ok {
// if it's a new entry, observe it and exit.
m.sameIP.WithLabelValues(chainID, version).Observe(1)
m.heartbeats.WithLabelValues(chainID, version).Inc()
return
}
if now.Sub(entry.Time) < MinHeartbeatInterval {
// if the span is still going, then add it up
atomic.AddUint64(&entry.Count, 1)
} else {
// if the span ended, then meter it, and reset it
m.sameIP.WithLabelValues(chainID, version).Observe(float64(atomic.LoadUint64(&entry.Count)))
entry.Time = now
atomic.StoreUint64(&entry.Count, 1)
m.heartbeats.WithLabelValues(chainID, version).Inc()
}
// always add, to keep LRU accurate
m.heartbeatUsers.Add(key, entry)
}
func (m *metrics) RecordVersion(version string) {
m.version.WithLabelValues(version).Set(1)
}