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
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
|
package cmd
import (
"encoding/json"
"fmt"
"io"
"math"
"os"
"time"
"github.com/spf13/cobra"
"github.com/vespa-engine/vespa/client/go/internal/vespa/document"
)
func addFeedFlags(cmd *cobra.Command, concurrency *int) {
// TOOD(mpolden): Remove this flag
cmd.PersistentFlags().IntVarP(concurrency, "concurrency", "T", 64, "Number of goroutines to use for dispatching")
}
func newFeedCmd(cli *CLI) *cobra.Command {
var (
concurrency int
)
cmd := &cobra.Command{
Use: "feed FILE",
Short: "Feed documents to a Vespa cluster",
Long: `Feed documents to a Vespa cluster.
A high performance feeding client. This can be used to feed large amounts of
documents to Vespa cluster efficiently.
The contents of FILE must be either a JSON array or JSON objects separated by
newline (JSONL).
`,
Example: `$ vespa feed documents.jsonl
`,
Args: cobra.ExactArgs(1),
DisableAutoGenTag: true,
SilenceUsage: true,
Hidden: true, // TODO(mpolden): Remove when ready for public use
RunE: func(cmd *cobra.Command, args []string) error {
f, err := os.Open(args[0])
if err != nil {
return err
}
defer f.Close()
return feed(f, cli, concurrency)
},
}
addFeedFlags(cmd, &concurrency)
return cmd
}
func feed(r io.Reader, cli *CLI, concurrency int) error {
service, err := documentService(cli)
if err != nil {
return err
}
client := document.NewClient(document.ClientOptions{
BaseURL: service.BaseURL,
}, service)
throttler := document.NewThrottler()
// TODO(mpolden): Make doom duration configurable
circuitBreaker := document.NewCircuitBreaker(10*time.Second, 0)
dispatcher := document.NewDispatcher(client, throttler, circuitBreaker)
dec := document.NewDecoder(r)
start := cli.now()
for {
doc, err := dec.Decode()
if err == io.EOF {
break
}
if err != nil {
cli.printErr(fmt.Errorf("failed to decode document: %w", err))
}
if err := dispatcher.Enqueue(doc); err != nil {
cli.printErr(err)
}
}
if err := dispatcher.Close(); err != nil {
return err
}
elapsed := cli.now().Sub(start)
return writeSummaryJSON(cli.Stdout, dispatcher.Stats(), elapsed)
}
type number float32
func (n number) MarshalJSON() ([]byte, error) { return []byte(fmt.Sprintf("%.3f", n)), nil }
type feedSummary struct {
Seconds number `json:"feeder.seconds"`
SuccessCount int64 `json:"feeder.ok.count"`
SuccessRate number `json:"feeder.ok.rate"`
ErrorCount int64 `json:"feeder.error.count"`
InflightCount int64 `json:"feeder.inflight.count"`
RequestCount int64 `json:"http.request.count"`
RequestBytes int64 `json:"http.request.bytes"`
RequestRate number `json:"http.request.MBps"`
ExceptionCount int64 `json:"http.exception.count"` // same as ErrorCount, for compatability with vespa-feed-client
ResponseCount int64 `json:"http.response.count"`
ResponseBytes int64 `json:"http.response.bytes"`
ResponseRate number `json:"http.response.MBps"`
ResponseErrorCount int64 `json:"http.response.error.count"`
ResponseMinLatency int64 `json:"http.response.latency.millis.min"`
ResponseAvgLatency int64 `json:"http.response.latency.millis.avg"`
ResponseMaxLatency int64 `json:"http.response.latency.millis.max"`
ResponseCodeCounts map[int]int64 `json:"http.response.code.counts"`
}
func mbps(bytes int64, duration time.Duration) float64 {
return (float64(bytes) / 1000 / 1000) / math.Max(1, duration.Seconds())
}
func writeSummaryJSON(w io.Writer, stats document.Stats, duration time.Duration) error {
summary := feedSummary{
Seconds: number(duration.Seconds()),
SuccessCount: stats.Successes(),
SuccessRate: number(float64(stats.Successes()) / math.Max(1, duration.Seconds())),
ErrorCount: stats.Errors,
InflightCount: stats.Inflight,
RequestCount: stats.Requests,
RequestBytes: stats.BytesSent,
RequestRate: number(mbps(stats.BytesSent, duration)),
ExceptionCount: stats.Errors,
ResponseCount: stats.Responses,
ResponseBytes: stats.BytesRecv,
ResponseRate: number(mbps(stats.BytesRecv, duration)),
ResponseErrorCount: stats.Responses - stats.Successes(),
ResponseMinLatency: stats.MinLatency.Milliseconds(),
ResponseAvgLatency: stats.AvgLatency().Milliseconds(),
ResponseMaxLatency: stats.MaxLatency.Milliseconds(),
ResponseCodeCounts: stats.ResponsesByCode,
}
enc := json.NewEncoder(w)
enc.SetIndent("", " ")
return enc.Encode(summary)
}
|