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
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
|
package logr
import (
"context"
"errors"
"fmt"
"os"
"sync/atomic"
"time"
)
// Target represents a destination for log records such as file,
// database, TCP socket, etc.
type Target interface {
// Init is called once to initialize the target.
Init() error
// Write outputs to this target's destination.
Write(p []byte, rec *LogRec) (int, error)
// Shutdown is called once to free/close any resources.
// Target queue is already drained when this is called.
Shutdown() error
}
type targetMetrics struct {
queueSizeGauge Gauge
loggedCounter Counter
errorCounter Counter
droppedCounter Counter
blockedCounter Counter
}
type targetHostOptions struct {
name string
filter Filter
formatter Formatter
maxQueueSize int
metrics *metrics
}
// TargetHost hosts and manages the lifecycle of a target.
// Incoming log records are queued and formatted before
// being passed to the target.
type TargetHost struct {
target Target
name string
filter Filter
formatter Formatter
in chan *LogRec
quit chan struct{} // closed by Shutdown to exit read loop
done chan struct{} // closed when read loop exited
targetMetrics *targetMetrics
shutdown int32
}
func newTargetHost(target Target, options targetHostOptions) (*TargetHost, error) {
host := &TargetHost{
target: target,
name: options.name,
filter: options.filter,
formatter: options.formatter,
in: make(chan *LogRec, options.maxQueueSize),
quit: make(chan struct{}),
done: make(chan struct{}),
}
if host.name == "" {
host.name = fmt.Sprintf("%T", target)
}
if host.filter == nil {
host.filter = &StdFilter{Lvl: Fatal}
}
if host.formatter == nil {
host.formatter = &DefaultFormatter{}
}
err := host.initMetrics(options.metrics)
if err != nil {
return nil, err
}
err = target.Init()
if err != nil {
return nil, err
}
go host.start()
return host, nil
}
func (h *TargetHost) initMetrics(metrics *metrics) error {
if metrics == nil {
return nil
}
var err error
tmetrics := &targetMetrics{}
if tmetrics.queueSizeGauge, err = metrics.collector.QueueSizeGauge(h.name); err != nil {
return err
}
if tmetrics.loggedCounter, err = metrics.collector.LoggedCounter(h.name); err != nil {
return err
}
if tmetrics.errorCounter, err = metrics.collector.ErrorCounter(h.name); err != nil {
return err
}
if tmetrics.droppedCounter, err = metrics.collector.DroppedCounter(h.name); err != nil {
return err
}
if tmetrics.blockedCounter, err = metrics.collector.BlockedCounter(h.name); err != nil {
return err
}
h.targetMetrics = tmetrics
updateFreqMillis := metrics.updateFreqMillis
if updateFreqMillis == 0 {
updateFreqMillis = DefMetricsUpdateFreqMillis
}
if updateFreqMillis < 250 {
updateFreqMillis = 250 // don't peg the CPU
}
go h.startMetricsUpdater(updateFreqMillis)
return nil
}
// IsLevelEnabled returns true if this target should emit logs for the specified level.
func (h *TargetHost) IsLevelEnabled(lvl Level) (enabled bool, level Level) {
level, enabled = h.filter.GetEnabledLevel(lvl)
return enabled, level
}
// Shutdown stops processing log records after making best
// effort to flush queue.
func (h *TargetHost) Shutdown(ctx context.Context) error {
if atomic.SwapInt32(&h.shutdown, 1) != 0 {
return errors.New("targetHost shutdown called more than once")
}
close(h.quit)
// No more records can be accepted; now wait for read loop to exit.
select {
case <-ctx.Done():
case <-h.done:
}
// b.in channel should now be drained.
return h.target.Shutdown()
}
// Log queues a log record to be output to this target's destination.
func (h *TargetHost) Log(rec *LogRec) {
if atomic.LoadInt32(&h.shutdown) != 0 {
return
}
lgr := rec.Logger().Logr()
select {
case h.in <- rec:
default:
handler := lgr.options.onTargetQueueFull
if handler != nil && handler(h.target, rec, cap(h.in)) {
h.incDroppedCounter()
return // drop the record
}
h.incBlockedCounter()
select {
case <-time.After(lgr.options.enqueueTimeout):
lgr.ReportError(fmt.Errorf("target enqueue timeout for log rec [%v]", rec))
case h.in <- rec: // block until success or timeout
}
}
}
func (h *TargetHost) setQueueSizeGauge(val float64) {
if h.targetMetrics != nil {
h.targetMetrics.queueSizeGauge.Set(val)
}
}
func (h *TargetHost) incLoggedCounter() {
if h.targetMetrics != nil {
h.targetMetrics.loggedCounter.Inc()
}
}
func (h *TargetHost) incErrorCounter() {
if h.targetMetrics != nil {
h.targetMetrics.errorCounter.Inc()
}
}
func (h *TargetHost) incDroppedCounter() {
if h.targetMetrics != nil {
h.targetMetrics.droppedCounter.Inc()
}
}
func (h *TargetHost) incBlockedCounter() {
if h.targetMetrics != nil {
h.targetMetrics.blockedCounter.Inc()
}
}
// String returns a name for this target.
func (h *TargetHost) String() string {
return h.name
}
// start accepts log records via In channel and writes to the
// supplied target, until Done channel signaled.
func (h *TargetHost) start() {
defer func() {
if r := recover(); r != nil {
fmt.Fprintln(os.Stderr, "TargetHost.start -- ", r)
go h.start()
} else {
close(h.done)
}
}()
for {
var rec *LogRec
select {
case rec = <-h.in:
if rec.flush != nil {
h.flush(rec.flush)
} else {
err := h.writeRec(rec)
if err != nil {
h.incErrorCounter()
rec.Logger().Logr().ReportError(err)
} else {
h.incLoggedCounter()
}
}
case <-h.quit:
return
}
}
}
func (h *TargetHost) writeRec(rec *LogRec) error {
level, enabled := h.filter.GetEnabledLevel(rec.Level())
if !enabled {
// how did we get here?
return fmt.Errorf("level %s not enabled for target %s", rec.Level().Name, h.name)
}
buf := rec.logger.lgr.BorrowBuffer()
defer rec.logger.lgr.ReleaseBuffer(buf)
buf, err := h.formatter.Format(rec, level, buf)
if err != nil {
return err
}
_, err = h.target.Write(buf.Bytes(), rec)
return err
}
// startMetricsUpdater updates the metrics for any polled values every `updateFreqMillis` seconds until
// target is shut down.
func (h *TargetHost) startMetricsUpdater(updateFreqMillis int64) {
for {
select {
case <-h.done:
return
case <-time.After(time.Duration(updateFreqMillis) * time.Millisecond):
h.setQueueSizeGauge(float64(len(h.in)))
}
}
}
// flush drains the queue and notifies when done.
func (h *TargetHost) flush(done chan<- struct{}) {
for {
var rec *LogRec
var err error
select {
case rec = <-h.in:
// ignore any redundant flush records.
if rec.flush == nil {
err = h.writeRec(rec)
if err != nil {
h.incErrorCounter()
rec.Logger().Logr().ReportError(err)
}
}
default:
done <- struct{}{}
return
}
}
}
|