-
Notifications
You must be signed in to change notification settings - Fork 1.7k
/
Copy pathflb_engine.c
1260 lines (1064 loc) · 39.6 KB
/
flb_engine.c
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
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
/* -*- Mode: C; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
/* Fluent Bit
* ==========
* Copyright (C) 2015-2024 The Fluent Bit Authors
*
* 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.
*/
#include <math.h>
#include <stdio.h>
#include <stdlib.h>
#include <monkey/mk_core.h>
#include <fluent-bit/flb_bucket_queue.h>
#include <fluent-bit/flb_event_loop.h>
#include <fluent-bit/flb_time.h>
#include <fluent-bit/flb_info.h>
#include <fluent-bit/flb_bits.h>
#include <fluent-bit/flb_macros.h>
#include <fluent-bit/flb_pipe.h>
#include <fluent-bit/flb_custom.h>
#include <fluent-bit/flb_input.h>
#include <fluent-bit/flb_output.h>
#include <fluent-bit/flb_error.h>
#include <fluent-bit/flb_utils.h>
#include <fluent-bit/flb_config.h>
#include <fluent-bit/flb_engine.h>
#include <fluent-bit/flb_engine_dispatch.h>
#include <fluent-bit/flb_network.h>
#include <fluent-bit/flb_task.h>
#include <fluent-bit/flb_router.h>
#include <fluent-bit/flb_http_server.h>
#include <fluent-bit/flb_scheduler.h>
#include <fluent-bit/flb_parser.h>
#include <fluent-bit/flb_sosreport.h>
#include <fluent-bit/flb_storage.h>
#include <fluent-bit/flb_http_server.h>
#include <fluent-bit/flb_pack.h>
#include <fluent-bit/flb_metrics.h>
#include <fluent-bit/flb_version.h>
#include <fluent-bit/flb_upstream.h>
#include <fluent-bit/flb_downstream.h>
#include <fluent-bit/flb_ring_buffer.h>
#include <fluent-bit/flb_notification.h>
#include <fluent-bit/flb_simd.h>
#ifdef FLB_HAVE_METRICS
#include <fluent-bit/flb_metrics_exporter.h>
#endif
#ifdef FLB_HAVE_STREAM_PROCESSOR
#include <fluent-bit/stream_processor/flb_sp.h>
#endif
#ifdef FLB_HAVE_AWS_ERROR_REPORTER
#include <fluent-bit/aws/flb_aws_error_reporter.h>
extern struct flb_aws_error_reporter *error_reporter;
#endif
#include <ctraces/ctr_version.h>
static pthread_once_t local_thread_engine_evl_init = PTHREAD_ONCE_INIT;
FLB_TLS_DEFINE(struct mk_event_loop, flb_engine_evl);
static void flb_engine_evl_init_private()
{
FLB_TLS_INIT(flb_engine_evl);
}
void flb_engine_evl_init()
{
pthread_once(&local_thread_engine_evl_init, flb_engine_evl_init_private);
}
struct mk_event_loop *flb_engine_evl_get()
{
struct mk_event_loop *evl;
evl = FLB_TLS_GET(flb_engine_evl);
return evl;
}
void flb_engine_evl_set(struct mk_event_loop *evl)
{
FLB_TLS_SET(flb_engine_evl, evl);
}
int flb_engine_destroy_tasks(struct mk_list *tasks)
{
int c = 0;
struct mk_list *tmp;
struct mk_list *head;
struct flb_task *task;
mk_list_foreach_safe(head, tmp, tasks) {
task = mk_list_entry(head, struct flb_task, _head);
flb_task_destroy(task, FLB_FALSE);
c++;
}
return c;
}
void flb_engine_reschedule_retries(struct flb_config *config)
{
int ret;
struct mk_list *head;
struct mk_list *t_head;
struct mk_list *rt_head;
struct mk_list *tmp_task;
struct mk_list *tmp_retry_task;
struct flb_task *task;
struct flb_input_instance *ins;
struct flb_task_retry *retry;
/* Invalidate and reschedule all retry tasks to be retried immediately */
mk_list_foreach(head, &config->inputs) {
ins = mk_list_entry(head, struct flb_input_instance, _head);
mk_list_foreach_safe(t_head, tmp_task, &ins->tasks) {
task = mk_list_entry(t_head, struct flb_task, _head);
mk_list_foreach_safe(rt_head, tmp_retry_task, &task->retries) {
retry = mk_list_entry(rt_head, struct flb_task_retry, _head);
flb_sched_request_invalidate(config, retry);
ret = flb_sched_retry_now(config, retry);
if (ret == -1) {
/* Can't do much here, just continue on */
flb_warn("[engine] failed to immediately re-schedule retry=%p "
"for task %i. Err: %d", retry, task->id, flb_errno());
} else {
flb_debug("[engine] re-scheduled retry=%p for task %i",
retry, task->id);
}
}
}
}
}
int flb_engine_flush(struct flb_config *config,
struct flb_input_plugin *in_force)
{
struct flb_input_instance *in;
struct flb_input_plugin *p;
struct mk_list *head;
mk_list_foreach(head, &config->inputs) {
in = mk_list_entry(head, struct flb_input_instance, _head);
p = in->p;
if (in_force != NULL && p != in_force) {
continue;
}
flb_engine_dispatch(0, in, config);
}
return 0;
}
/* Cleanup function that runs every 1.5 second */
static void cb_engine_sched_timer(struct flb_config *ctx, void *data)
{
(void) data;
/* Upstream timeout handling */
flb_upstream_conn_timeouts(&ctx->upstreams);
/* Downstream timeout handling */
flb_downstream_conn_timeouts(&ctx->downstreams);
}
static inline int handle_input_event(flb_pipefd_t fd, uint64_t ts,
struct flb_config *config)
{
int bytes;
uint32_t type;
uint32_t ins_id;
uint64_t val;
bytes = flb_pipe_r(fd, &val, sizeof(val));
if (bytes == -1) {
flb_pipe_error();
return -1;
}
/* Get type and key */
type = FLB_BITS_U64_HIGH(val);
ins_id = FLB_BITS_U64_LOW(val);
/* At the moment we only support events coming from an input coroutine */
if (type != FLB_ENGINE_IN_CORO) {
flb_error("[engine] invalid event type %i for input handler",
type);
return -1;
}
flb_input_coro_finished(config, (int) ins_id);
return 0;
}
static inline double calculate_chunk_capacity_percent(struct flb_output_instance *ins)
{
/* Currently, total_limit_size 0(K|M)B will be translated as no
* limit. So, we need to handle this situation to be unlimited. */
if (ins->total_limit_size <= 0) {
return 100.0;
}
return 100 * (1.0 - (ins->fs_backlog_chunks_size + ins->fs_chunks_size)/
((double)ins->total_limit_size));
}
static inline int handle_output_event(uint64_t ts,
struct flb_config *config,
uint64_t val)
{
int ret;
int task_id;
int out_id;
int retries;
int retry_seconds;
uint32_t type;
uint32_t key;
char *name;
struct flb_task *task;
struct flb_task_retry *retry;
struct flb_output_instance *ins;
/* Get type and key */
type = FLB_BITS_U64_HIGH(val);
key = FLB_BITS_U64_LOW(val);
if (type != FLB_ENGINE_TASK) {
flb_error("[engine] invalid event type %i for output handler",
type);
return -1;
}
/*
* The notion of ENGINE_TASK is associated to outputs. All thread
* references below belongs to flb_output_coro's.
*/
ret = FLB_TASK_RET(key);
task_id = FLB_TASK_ID(key);
out_id = FLB_TASK_OUT(key);
#ifdef FLB_HAVE_TRACE
char *trace_st = NULL;
if (ret == FLB_OK) {
trace_st = "OK";
}
else if (ret == FLB_ERROR) {
trace_st = "ERROR";
}
else if (ret == FLB_RETRY) {
trace_st = "RETRY";
}
flb_trace("%s[engine] [task event]%s task_id=%i out_id=%i return=%s",
ANSI_YELLOW, ANSI_RESET,
task_id, out_id, trace_st);
#endif
task = config->task_map[task_id].task;
ins = flb_output_get_instance(config, out_id);
if (flb_output_is_threaded(ins) == FLB_FALSE) {
flb_output_flush_finished(config, out_id);
}
name = (char *) flb_output_name(ins);
/* If we are in synchronous mode, flush the next waiting task */
if (ins->flags & FLB_OUTPUT_SYNCHRONOUS) {
if (ret == FLB_OK || ret == FLB_RETRY || ret == FLB_ERROR) {
flb_output_task_singleplex_flush_next(ins->singleplex_queue);
}
}
/* A task has finished, delete it */
if (ret == FLB_OK) {
/* cmetrics */
cmt_counter_add(ins->cmt_proc_records, ts, task->event_chunk->total_events,
1, (char *[]) {name});
cmt_counter_add(ins->cmt_proc_bytes, ts, task->event_chunk->size,
1, (char *[]) {name});
/* [OLD API] Update metrics */
#ifdef FLB_HAVE_METRICS
if (ins->metrics) {
flb_metrics_sum(FLB_METRIC_OUT_OK_RECORDS,
task->event_chunk->total_events, ins->metrics);
flb_metrics_sum(FLB_METRIC_OUT_OK_BYTES,
task->event_chunk->size, ins->metrics);
}
#endif
/* Inform the user if a 'retry' succedeed */
if (mk_list_size(&task->retries) > 0) {
retries = flb_task_retry_count(task, ins);
if (retries > 0) {
flb_info("[engine] flush chunk '%s' succeeded at retry %i: "
"task_id=%i, input=%s > output=%s (out_id=%i)",
flb_input_chunk_get_name(task->ic),
retries, task_id,
flb_input_name(task->i_ins),
flb_output_name(ins), out_id);
}
}
else if (flb_task_from_fs_storage(task) == FLB_TRUE) {
flb_info("[engine] flush backlog chunk '%s' succeeded: "
"task_id=%i, input=%s > output=%s (out_id=%i)",
flb_input_chunk_get_name(task->ic),
task_id,
flb_input_name(task->i_ins),
flb_output_name(ins), out_id);
}
cmt_gauge_set(ins->cmt_chunk_available_capacity_percent, ts,
calculate_chunk_capacity_percent(ins),
1, (char *[]) {name});
flb_task_retry_clean(task, ins);
flb_task_users_dec(task, FLB_TRUE);
}
else if (ret == FLB_RETRY) {
if (ins->retry_limit == FLB_OUT_RETRY_NONE) {
/* cmetrics: output_dropped_records_total */
cmt_counter_add(ins->cmt_dropped_records, ts, task->records,
1, (char *[]) {name});
cmt_gauge_set(ins->cmt_chunk_available_capacity_percent, ts,
calculate_chunk_capacity_percent(ins),
1, (char *[]) {name});
/* OLD metrics API */
#ifdef FLB_HAVE_METRICS
flb_metrics_sum(FLB_METRIC_OUT_DROPPED_RECORDS, task->records, ins->metrics);
#endif
flb_info("[engine] chunk '%s' is not retried (no retry config): "
"task_id=%i, input=%s > output=%s (out_id=%i)",
flb_input_chunk_get_name(task->ic),
task_id,
flb_input_name(task->i_ins),
flb_output_name(ins), out_id);
flb_task_retry_clean(task, ins);
flb_task_users_dec(task, FLB_TRUE);
return 0;
}
/* Create a Task-Retry */
retry = flb_task_retry_create(task, ins);
if (!retry) {
/*
* It can fail in two situations:
*
* - No enough memory (unlikely)
* - It reached the maximum number of re-tries
*/
/* cmetrics */
cmt_counter_inc(ins->cmt_retries_failed, ts, 1, (char *[]) {name});
cmt_counter_add(ins->cmt_dropped_records, ts, task->records,
1, (char *[]) {name});
cmt_gauge_set(ins->cmt_chunk_available_capacity_percent, ts,
calculate_chunk_capacity_percent(ins),
1, (char *[]) {name});
/* OLD metrics API */
#ifdef FLB_HAVE_METRICS
flb_metrics_sum(FLB_METRIC_OUT_RETRY_FAILED, 1, ins->metrics);
flb_metrics_sum(FLB_METRIC_OUT_DROPPED_RECORDS, task->records, ins->metrics);
#endif
/* Notify about this failed retry */
flb_error("[engine] chunk '%s' cannot be retried: "
"task_id=%i, input=%s > output=%s",
flb_input_chunk_get_name(task->ic),
task_id,
flb_input_name(task->i_ins),
flb_output_name(ins));
flb_task_retry_clean(task, ins);
flb_task_users_dec(task, FLB_TRUE);
return 0;
}
/* Always destroy the old coroutine */
flb_task_users_dec(task, FLB_FALSE);
/* Let the scheduler to retry the failed task/thread */
retry_seconds = flb_sched_request_create(config,
retry, retry->attempts);
/*
* If for some reason the Scheduler could not include this retry,
* we need to get rid of it, likely this is because of not enough
* memory available or we ran out of file descriptors.
*/
if (retry_seconds == -1) {
flb_warn("[engine] retry for chunk '%s' could not be scheduled: "
"input=%s > output=%s",
flb_input_chunk_get_name(task->ic),
flb_input_name(task->i_ins),
flb_output_name(ins));
flb_task_retry_destroy(retry);
flb_task_users_release(task);
}
else {
/* Inform the user 'retry' has been scheduled */
flb_warn("[engine] failed to flush chunk '%s', retry in %i seconds: "
"task_id=%i, input=%s > output=%s (out_id=%i)",
flb_input_chunk_get_name(task->ic),
retry_seconds,
task->id,
flb_input_name(task->i_ins),
flb_output_name(ins), out_id);
/* cmetrics */
cmt_counter_inc(ins->cmt_retries, ts, 1, (char *[]) {name});
cmt_counter_add(ins->cmt_retried_records, ts, task->records,
1, (char *[]) {name});
cmt_gauge_set(ins->cmt_chunk_available_capacity_percent, ts,
calculate_chunk_capacity_percent(ins),
1, (char *[]) {name});
/* OLD metrics API: update the metrics since a new retry is coming */
#ifdef FLB_HAVE_METRICS
flb_metrics_sum(FLB_METRIC_OUT_RETRY, 1, ins->metrics);
flb_metrics_sum(FLB_METRIC_OUT_RETRIED_RECORDS, task->records, ins->metrics);
#endif
}
}
else if (ret == FLB_ERROR) {
/* cmetrics */
cmt_counter_inc(ins->cmt_errors, ts, 1, (char *[]) {name});
cmt_counter_add(ins->cmt_dropped_records, ts, task->records,
1, (char *[]) {name});
cmt_gauge_set(ins->cmt_chunk_available_capacity_percent, ts,
calculate_chunk_capacity_percent(ins),
1, (char *[]) {name});
/* OLD API */
#ifdef FLB_HAVE_METRICS
flb_metrics_sum(FLB_METRIC_OUT_ERROR, 1, ins->metrics);
flb_metrics_sum(FLB_METRIC_OUT_DROPPED_RECORDS, task->records, ins->metrics);
#endif
flb_task_retry_clean(task, ins);
flb_task_users_dec(task, FLB_TRUE);
}
return 0;
}
static inline int handle_output_events(flb_pipefd_t fd,
struct flb_config *config)
{
uint64_t values[FLB_ENGINE_OUTPUT_EVENT_BATCH_SIZE];
int result;
int bytes;
size_t limit;
size_t index;
uint64_t ts;
memset(&values, 0, sizeof(values));
bytes = flb_pipe_r(fd, &values, sizeof(values));
if (bytes == -1) {
flb_pipe_error();
return -1;
}
limit = floor(bytes / sizeof(uint64_t));
ts = cfl_time_now();
for (index = 0 ;
index < limit &&
index < (sizeof(values) / sizeof(values[0])) ;
index++) {
if (values[index] == 0) {
break;
}
result = handle_output_event(ts, config, values[index]);
}
/* This is wrong, in one hand, if handle_output_event_ fails we should
* stop, on the other, we have already consumed the signals from the pipe
* so we have to do whatever we can with them.
*
* And a side effect is that since we have N results but we are not aborting
* as soon as we get an error there could be N results to this function which
* not only are we not ready to handle but is not even checked at the moment.
*/
return result;
}
static inline int flb_engine_manager(flb_pipefd_t fd, struct flb_config *config)
{
int bytes;
uint32_t type;
uint32_t key;
uint64_t val;
/* read the event */
bytes = flb_pipe_r(fd, &val, sizeof(val));
if (bytes == -1) {
flb_pipe_error();
return -1;
}
/* Get type and key */
type = FLB_BITS_U64_HIGH(val);
key = FLB_BITS_U64_LOW(val);
/* Flush all remaining data */
if (type == 1) { /* Engine type */
if (key == FLB_ENGINE_STOP) {
flb_trace("[engine] flush enqueued data");
flb_engine_flush(config, NULL);
return FLB_ENGINE_STOP;
}
}
return 0;
}
static FLB_INLINE int flb_engine_handle_event(flb_pipefd_t fd, int mask,
struct flb_config *config)
{
int64_t ret;
/* flb_engine_shutdown was already initiated */
if (config->is_running == FLB_FALSE) {
return 0;
}
if (mask & MK_EVENT_READ) {
/* Check if we need to flush */
if (config->flush_fd == fd) {
flb_utils_timer_consume(fd);
flb_engine_flush(config, NULL);
return 0;
}
else if (config->shutdown_fd == fd) {
flb_utils_pipe_byte_consume(fd);
return FLB_ENGINE_SHUTDOWN;
}
else if (config->ch_manager[0] == fd) {
ret = flb_engine_manager(fd, config);
if (ret == FLB_ENGINE_STOP || ret == FLB_ENGINE_EV_STOP) {
return FLB_ENGINE_STOP;
}
}
/* Try to match the file descriptor with a collector event */
ret = flb_input_collector_fd(fd, config);
if (ret != -1) {
return ret;
}
/* Metrics exporter event ? */
#ifdef FLB_HAVE_METRICS
ret = flb_me_fd_event(fd, config->metrics);
if (ret != -1) {
return ret;
}
#endif
/* Stream processor event ? */
#ifdef FLB_HAVE_STREAM_PROCESSOR
if (config->stream_processor_ctx) {
ret = flb_sp_fd_event(fd, config->stream_processor_ctx);
if (ret != -1) {
return ret;
}
}
#endif
}
return 0;
}
static int flb_engine_started(struct flb_config *config)
{
uint64_t val;
/* Check the channel is valid (enabled by library mode) */
if (config->ch_notif[1] <= 0) {
return -1;
}
val = FLB_ENGINE_STARTED;
return flb_pipe_w(config->ch_notif[1], &val, sizeof(uint64_t));
}
int flb_engine_failed(struct flb_config *config)
{
int ret;
uint64_t val;
/* Check the channel is valid (enabled by library mode) */
if (config->ch_notif[1] <= 0) {
flb_error("[engine] no channel to notify FAILED message");
return -1;
}
val = FLB_ENGINE_FAILED;
ret = flb_pipe_w(config->ch_notif[1], &val, sizeof(uint64_t));
if (ret == -1) {
flb_error("[engine] fail to dispatch FAILED message");
}
/* Waiting flushing log */
sleep(1);
return ret;
}
static int flb_engine_log_start(struct flb_config *config)
{
int type;
int level;
/* Log Level */
if (config->verbose != FLB_LOG_INFO) {
level = config->verbose;
}
else {
level = FLB_LOG_INFO;
}
/* Destination based on type */
if (config->log_file) {
type = FLB_LOG_FILE;
}
else {
type = FLB_LOG_STDERR;
}
if (flb_log_create(config, type, level, config->log_file) == NULL) {
return -1;
}
return 0;
}
static void flb_engine_drain_ring_buffer_signal_channel(flb_pipefd_t fd)
{
static char signal_buffer[512];
flb_pipe_r(fd, signal_buffer, sizeof(signal_buffer));
}
#ifdef FLB_HAVE_IN_STORAGE_BACKLOG
extern int sb_segregate_chunks(struct flb_config *config);
#else
int sb_segregate_chunks(struct flb_config *config)
{
return 0;
}
#endif
int flb_engine_start(struct flb_config *config)
{
int ret;
uint64_t ts;
char tmp[16];
int rb_flush_flag;
struct flb_time t_flush;
struct mk_event *event;
struct mk_event_loop *evl;
struct flb_bucket_queue *evl_bktq;
struct flb_sched *sched;
struct flb_net_dns dns_ctx;
struct flb_notification *notification;
enum ctx_signal_type ctx_signal;
/* Initialize the networking layer */
flb_net_lib_init();
flb_net_ctx_init(&dns_ctx);
flb_net_dns_ctx_init();
flb_net_dns_ctx_set(&dns_ctx);
flb_pack_init(config);
/* Create the event loop and set it in the global configuration */
evl = mk_event_loop_create(256);
if (!evl) {
fprintf(stderr, "[log] could not create event loop\n");
return -1;
}
config->evl = evl;
/* Create the bucket queue (FLB_ENGINE_PRIORITY_COUNT priorities) */
evl_bktq = flb_bucket_queue_create(FLB_ENGINE_PRIORITY_COUNT);
if (!evl_bktq) {
return -1;
}
config->evl_bktq = evl_bktq;
/*
* Event loop channel to send context signals.
*
*/
/* Create the event loop and set it in the global configuration */
config->ctx_evl = mk_event_loop_create(8);
if (!config->ctx_evl) {
fprintf(stderr, "[log] could not create context event loop\n");
return -1;
}
ret = mk_event_channel_create(config->ctx_evl,
&config->ch_context_signal[0],
&config->ch_context_signal[1],
&config->event_context_signal);
if (ret == -1) {
flb_error("[engine] could not create context signal channel");
return -1;
}
/* Signal type to indicate a "context" event */
config->event_context_signal.type = FLB_CONTEXT_EV_SIGNAL;
config->event_context_signal.priority = FLB_ENGINE_PRIORITY_THREAD;
/*
* Event loop channel to ingest flush events from flb_engine_flush()
*
* - FLB engine uses 'ch_self_events[1]' to dispatch tasks to self
* - Self to receive message on ch_parent_events[0]
*
* The mk_event_channel_create() will attach the pipe read end ch_self_events[0]
* to the local event loop 'evl'.
*/
ret = mk_event_channel_create(config->evl,
&config->ch_self_events[0],
&config->ch_self_events[1],
&config->event_thread_init);
if (ret == -1) {
flb_error("[engine] could not create engine thread channel");
return -1;
}
/* Signal type to indicate a "flush" request */
config->event_thread_init.type = FLB_ENGINE_EV_THREAD_ENGINE;
config->event_thread_init.priority = FLB_ENGINE_PRIORITY_THREAD;
/* Register the event loop on this thread */
flb_engine_evl_init();
flb_engine_evl_set(evl);
/* Start the Logging service */
ret = flb_engine_log_start(config);
if (ret == -1) {
fprintf(stderr, "[engine] log start failed\n");
return -1;
}
flb_info("[fluent bit] version=%s, commit=%.10s, pid=%i",
FLB_VERSION_STR, FLB_GIT_HASH, getpid());
/* Debug coroutine stack size */
flb_utils_bytes_to_human_readable_size(config->coro_stack_size,
tmp, sizeof(tmp));
flb_debug("[engine] coroutine stack size: %u bytes (%s)",
config->coro_stack_size, tmp);
/*
* Create a communication channel: this routine creates a channel to
* signal the Engine event loop. It's useful to stop the event loop
* or to instruct anything else without break.
*/
ret = mk_event_channel_create(config->evl,
&config->ch_manager[0],
&config->ch_manager[1],
&config->ch_event);
if (ret != 0) {
flb_error("[engine] could not create manager channels");
return -1;
}
ret = mk_event_channel_create(config->evl,
&config->notification_channels[0],
&config->notification_channels[1],
&config->notification_event);
if (ret == -1) {
flb_error("could not create main notification channel");
return -1;
}
config->notification_channels_initialized = FLB_TRUE;
config->notification_event.type = FLB_ENGINE_EV_NOTIFICATION;
ret = flb_routes_mask_set_size(mk_list_size(&config->outputs), config);
if (ret != 0) {
flb_error("[engine] routing mask dimensioning failed");
return -1;
}
ret = flb_routes_mask_set_size(mk_list_size(&config->outputs), config);
if (ret != 0) {
flb_error("[engine] routing mask dimensioning failed");
return -1;
}
/* Initialize custom plugins */
ret = flb_custom_init_all(config);
if (ret == -1) {
return -1;
}
/* Start the Storage engine */
ret = flb_storage_create(config);
if (ret == -1) {
flb_error("[engine] storage creation failed");
return -1;
}
/* Internals */
flb_info("[simd ] %s", flb_simd_info());
/* Init Metrics engine */
cmt_initialize();
flb_info("[cmetrics] version=%s", cmt_version());
flb_info("[ctraces ] version=%s", ctr_version());
/* Initialize the scheduler */
sched = flb_sched_create(config, config->evl);
if (!sched) {
flb_error("[engine] scheduler could not start");
return -1;
}
config->sched = sched;
/* Register the scheduler context */
flb_sched_ctx_init();
flb_sched_ctx_set(sched);
/* Initialize input plugins */
ret = flb_input_init_all(config);
if (ret == -1) {
flb_error("[engine] input initialization failed");
return -1;
}
/* Initialize filter plugins */
ret = flb_filter_init_all(config);
if (ret == -1) {
flb_error("[engine] filter initialization failed");
return -1;
}
/* Inputs pre-run */
flb_input_pre_run_all(config);
/* Initialize output plugins */
ret = flb_output_init_all(config);
if (ret == -1) {
flb_error("[engine] output initialization failed");
return -1;
}
/* Outputs pre-run */
flb_output_pre_run(config);
/* Create and register the timer fd for flush procedure */
event = &config->event_flush;
event->mask = MK_EVENT_EMPTY;
event->status = MK_EVENT_NONE;
flb_time_from_double(&t_flush, config->flush);
config->flush_fd = mk_event_timeout_create(evl,
t_flush.tm.tv_sec,
t_flush.tm.tv_nsec,
event);
event->priority = FLB_ENGINE_PRIORITY_FLUSH;
if (config->flush_fd == -1) {
flb_utils_error(FLB_ERR_CFG_FLUSH_CREATE);
}
#ifdef FLB_HAVE_METRICS
if (config->storage_metrics == FLB_TRUE) {
config->storage_metrics_ctx = flb_storage_metrics_create(config);
}
#endif
/* Prepare routing paths */
ret = flb_router_io_set(config);
if (ret == -1) {
flb_error("[engine] router failed");
return -1;
}
/* Support mode only */
if (config->support_mode == FLB_TRUE) {
sleep(1);
flb_sosreport(config);
exit(1);
}
/* Initialize Metrics exporter */
#ifdef FLB_HAVE_METRICS
config->metrics = flb_me_create(config);
#endif
/* Initialize HTTP Server */
#ifdef FLB_HAVE_HTTP_SERVER
if (config->http_server == FLB_TRUE) {
config->http_ctx = flb_hs_create(config->http_listen, config->http_port,
config);
flb_hs_start(config->http_ctx);
}
#endif
#ifdef FLB_HAVE_STREAM_PROCESSOR
config->stream_processor_ctx = flb_sp_create(config);
if (!config->stream_processor_ctx) {
flb_error("[engine] could not initialize stream processor");
}
#endif
/* Initialize collectors */
flb_input_collectors_start(config);
/*
* Sched a permanent callback triggered every 1.5 second to let other
* Fluent Bit components run tasks at that interval.
*/
ret = flb_sched_timer_cb_create(config->sched,
FLB_SCHED_TIMER_CB_PERM,
1500, cb_engine_sched_timer, config, NULL);
if (ret == -1) {
flb_error("[engine] could not schedule permanent callback");
return -1;
}
/* DEV/TEST change only */
int rb_ms;
char *rb_env;
rb_env = getenv("FLB_DEV_RB_MS");
if (!rb_env) {
rb_ms = 250;
}
else {
rb_ms = atoi(rb_env);
}
/* Input instance / Ring buffer collector */
ret = flb_sched_timer_cb_create(config->sched,
FLB_SCHED_TIMER_CB_PERM,
rb_ms, flb_input_chunk_ring_buffer_collector,
config, NULL);
if (ret == -1) {
flb_error("[engine] could not schedule permanent callback");
return -1;
}
/* Signal that we have started */
flb_engine_started(config);
ret = sb_segregate_chunks(config);
if (ret) {
flb_error("[engine] could not segregate backlog chunks");
return -2;
}
while (1) {
rb_flush_flag = FLB_FALSE;
mk_event_wait(evl); /* potentially conditional mk_event_wait or mk_event_wait_2 based on bucket queue capacity for one shot events */
flb_event_priority_live_foreach(event, evl_bktq, evl, FLB_ENGINE_LOOP_MAX_ITER) {
if (event->type == FLB_ENGINE_EV_CORE) {
ret = flb_engine_handle_event(event->fd, event->mask, config);
if (ret == FLB_ENGINE_STOP) {
if (config->grace_count == 0) {