forked from netdata/netdata
-
Notifications
You must be signed in to change notification settings - Fork 0
/
proc_diskstats.c
1700 lines (1386 loc) · 74.7 KB
/
proc_diskstats.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
// SPDX-License-Identifier: GPL-3.0-or-later
#include "plugin_proc.h"
#define RRD_TYPE_DISK "disk"
#define PLUGIN_PROC_MODULE_DISKSTATS_NAME "/proc/diskstats"
#define CONFIG_SECTION_PLUGIN_PROC_DISKSTATS "plugin:" PLUGIN_PROC_CONFIG_NAME ":" PLUGIN_PROC_MODULE_DISKSTATS_NAME
#define DISK_TYPE_UNKNOWN 0
#define DISK_TYPE_PHYSICAL 1
#define DISK_TYPE_PARTITION 2
#define DISK_TYPE_VIRTUAL 3
#define DEFAULT_PREFERRED_IDS "*"
#define DEFAULT_EXCLUDED_DISKS "loop* ram*"
static struct disk {
char *disk; // the name of the disk (sda, sdb, etc, after being looked up)
char *device; // the device of the disk (before being looked up)
unsigned long major;
unsigned long minor;
int sector_size;
int type;
char *mount_point;
// disk options caching
int do_io;
int do_ops;
int do_mops;
int do_iotime;
int do_qops;
int do_util;
int do_backlog;
int do_bcache;
int updated;
int device_is_bcache;
char *bcache_filename_dirty_data;
char *bcache_filename_writeback_rate;
char *bcache_filename_cache_congested;
char *bcache_filename_cache_available_percent;
char *bcache_filename_stats_five_minute_cache_hit_ratio;
char *bcache_filename_stats_hour_cache_hit_ratio;
char *bcache_filename_stats_day_cache_hit_ratio;
char *bcache_filename_stats_total_cache_hit_ratio;
char *bcache_filename_stats_total_cache_hits;
char *bcache_filename_stats_total_cache_misses;
char *bcache_filename_stats_total_cache_miss_collisions;
char *bcache_filename_stats_total_cache_bypass_hits;
char *bcache_filename_stats_total_cache_bypass_misses;
char *bcache_filename_stats_total_cache_readaheads;
char *bcache_filename_cache_read_races;
char *bcache_filename_cache_io_errors;
char *bcache_filename_priority_stats;
usec_t bcache_priority_stats_update_every_usec;
usec_t bcache_priority_stats_elapsed_usec;
RRDSET *st_io;
RRDDIM *rd_io_reads;
RRDDIM *rd_io_writes;
RRDSET *st_ops;
RRDDIM *rd_ops_reads;
RRDDIM *rd_ops_writes;
RRDSET *st_qops;
RRDDIM *rd_qops_operations;
RRDSET *st_backlog;
RRDDIM *rd_backlog_backlog;
RRDSET *st_util;
RRDDIM *rd_util_utilization;
RRDSET *st_mops;
RRDDIM *rd_mops_reads;
RRDDIM *rd_mops_writes;
RRDSET *st_iotime;
RRDDIM *rd_iotime_reads;
RRDDIM *rd_iotime_writes;
RRDSET *st_await;
RRDDIM *rd_await_reads;
RRDDIM *rd_await_writes;
RRDSET *st_avgsz;
RRDDIM *rd_avgsz_reads;
RRDDIM *rd_avgsz_writes;
RRDSET *st_svctm;
RRDDIM *rd_svctm_svctm;
RRDSET *st_bcache_size;
RRDDIM *rd_bcache_dirty_size;
RRDSET *st_bcache_usage;
RRDDIM *rd_bcache_available_percent;
RRDSET *st_bcache_hit_ratio;
RRDDIM *rd_bcache_hit_ratio_5min;
RRDDIM *rd_bcache_hit_ratio_1hour;
RRDDIM *rd_bcache_hit_ratio_1day;
RRDDIM *rd_bcache_hit_ratio_total;
RRDSET *st_bcache;
RRDDIM *rd_bcache_hits;
RRDDIM *rd_bcache_misses;
RRDDIM *rd_bcache_miss_collisions;
RRDSET *st_bcache_bypass;
RRDDIM *rd_bcache_bypass_hits;
RRDDIM *rd_bcache_bypass_misses;
RRDSET *st_bcache_rates;
RRDDIM *rd_bcache_rate_congested;
RRDDIM *rd_bcache_readaheads;
RRDDIM *rd_bcache_rate_writeback;
RRDSET *st_bcache_cache_allocations;
RRDDIM *rd_bcache_cache_allocations_unused;
RRDDIM *rd_bcache_cache_allocations_clean;
RRDDIM *rd_bcache_cache_allocations_dirty;
RRDDIM *rd_bcache_cache_allocations_metadata;
RRDDIM *rd_bcache_cache_allocations_unknown;
RRDSET *st_bcache_cache_read_races;
RRDDIM *rd_bcache_cache_read_races;
RRDDIM *rd_bcache_cache_io_errors;
struct disk *next;
} *disk_root = NULL;
#define rrdset_obsolete_and_pointer_null(st) do { if(st) { rrdset_is_obsolete(st); (st) = NULL; } } while(st)
// static char *path_to_get_hw_sector_size = NULL;
// static char *path_to_get_hw_sector_size_partitions = NULL;
static char *path_to_sys_dev_block_major_minor_string = NULL;
static char *path_to_sys_block_device = NULL;
static char *path_to_sys_block_device_bcache = NULL;
static char *path_to_sys_devices_virtual_block_device = NULL;
static char *path_to_device_mapper = NULL;
static char *path_to_device_label = NULL;
static char *path_to_device_id = NULL;
static char *path_to_veritas_volume_groups = NULL;
static int name_disks_by_id = CONFIG_BOOLEAN_NO;
static int global_bcache_priority_stats_update_every = 0; // disabled by default
static int global_enable_new_disks_detected_at_runtime = CONFIG_BOOLEAN_YES,
global_enable_performance_for_physical_disks = CONFIG_BOOLEAN_AUTO,
global_enable_performance_for_virtual_disks = CONFIG_BOOLEAN_AUTO,
global_enable_performance_for_partitions = CONFIG_BOOLEAN_NO,
global_do_io = CONFIG_BOOLEAN_AUTO,
global_do_ops = CONFIG_BOOLEAN_AUTO,
global_do_mops = CONFIG_BOOLEAN_AUTO,
global_do_iotime = CONFIG_BOOLEAN_AUTO,
global_do_qops = CONFIG_BOOLEAN_AUTO,
global_do_util = CONFIG_BOOLEAN_AUTO,
global_do_backlog = CONFIG_BOOLEAN_AUTO,
global_do_bcache = CONFIG_BOOLEAN_AUTO,
globals_initialized = 0,
global_cleanup_removed_disks = 1;
static SIMPLE_PATTERN *preferred_ids = NULL;
static SIMPLE_PATTERN *excluded_disks = NULL;
static unsigned long long int bcache_read_number_with_units(const char *filename) {
char buffer[50 + 1];
if(read_file(filename, buffer, 50) == 0) {
static int unknown_units_error = 10;
char *end = NULL;
long double value = str2ld(buffer, &end);
if(end && *end) {
if(*end == 'k')
return (unsigned long long int)(value * 1024.0);
else if(*end == 'M')
return (unsigned long long int)(value * 1024.0 * 1024.0);
else if(*end == 'G')
return (unsigned long long int)(value * 1024.0 * 1024.0 * 1024.0);
else if(*end == 'T')
return (unsigned long long int)(value * 1024.0 * 1024.0 * 1024.0 * 1024.0);
else if(unknown_units_error > 0) {
error("bcache file '%s' provides value '%s' with unknown units '%s'", filename, buffer, end);
unknown_units_error--;
}
}
return (unsigned long long int)value;
}
return 0;
}
void bcache_read_priority_stats(struct disk *d, const char *family, int update_every, usec_t dt) {
static procfile *ff = NULL;
static char *separators = " \t:%[]";
static ARL_BASE *arl_base = NULL;
static unsigned long long unused;
static unsigned long long clean;
static unsigned long long dirty;
static unsigned long long metadata;
static unsigned long long unknown;
// check if it is time to update this metric
d->bcache_priority_stats_elapsed_usec += dt;
if(likely(d->bcache_priority_stats_elapsed_usec < d->bcache_priority_stats_update_every_usec)) return;
d->bcache_priority_stats_elapsed_usec = 0;
// initialize ARL
if(unlikely(!arl_base)) {
arl_base = arl_create("bcache/priority_stats", NULL, 60);
arl_expect(arl_base, "Unused", &unused);
arl_expect(arl_base, "Clean", &clean);
arl_expect(arl_base, "Dirty", &dirty);
arl_expect(arl_base, "Metadata", &metadata);
}
ff = procfile_reopen(ff, d->bcache_filename_priority_stats, separators, PROCFILE_FLAG_DEFAULT);
if(likely(ff)) ff = procfile_readall(ff);
if(unlikely(!ff)) {
separators = " \t:%[]";
return;
}
// do not reset the separators on every iteration
separators = NULL;
arl_begin(arl_base);
unused = clean = dirty = metadata = unknown = 0;
size_t lines = procfile_lines(ff), l;
for(l = 0; l < lines ;l++) {
size_t words = procfile_linewords(ff, l);
if(unlikely(words < 2)) {
if(unlikely(words)) error("Cannot read '%s' line %zu. Expected 2 params, read %zu.", d->bcache_filename_priority_stats, l, words);
continue;
}
if(unlikely(arl_check(arl_base,
procfile_lineword(ff, l, 0),
procfile_lineword(ff, l, 1)))) break;
}
unknown = 100 - unused - clean - dirty - metadata;
// create / update the cache allocations chart
{
if(unlikely(!d->st_bcache_cache_allocations)) {
d->st_bcache_cache_allocations = rrdset_create_localhost(
"disk_bcache_cache_alloc"
, d->device
, d->disk
, family
, "disk.bcache_cache_alloc"
, "BCache Cache Allocations"
, "percentage"
, PLUGIN_PROC_NAME
, PLUGIN_PROC_MODULE_DISKSTATS_NAME
, NETDATA_CHART_PRIO_BCACHE_CACHE_ALLOC
, update_every
, RRDSET_TYPE_STACKED
);
d->rd_bcache_cache_allocations_unused = rrddim_add(d->st_bcache_cache_allocations, "unused", NULL, 1, 1, RRD_ALGORITHM_ABSOLUTE);
d->rd_bcache_cache_allocations_dirty = rrddim_add(d->st_bcache_cache_allocations, "dirty", NULL, 1, 1, RRD_ALGORITHM_ABSOLUTE);
d->rd_bcache_cache_allocations_clean = rrddim_add(d->st_bcache_cache_allocations, "clean", NULL, 1, 1, RRD_ALGORITHM_ABSOLUTE);
d->rd_bcache_cache_allocations_metadata = rrddim_add(d->st_bcache_cache_allocations, "metadata", NULL, 1, 1, RRD_ALGORITHM_ABSOLUTE);
d->rd_bcache_cache_allocations_unknown = rrddim_add(d->st_bcache_cache_allocations, "undefined", NULL, 1, 1, RRD_ALGORITHM_ABSOLUTE);
d->bcache_priority_stats_update_every_usec = update_every * USEC_PER_SEC;
}
else rrdset_next(d->st_bcache_cache_allocations);
rrddim_set_by_pointer(d->st_bcache_cache_allocations, d->rd_bcache_cache_allocations_unused, unused);
rrddim_set_by_pointer(d->st_bcache_cache_allocations, d->rd_bcache_cache_allocations_dirty, dirty);
rrddim_set_by_pointer(d->st_bcache_cache_allocations, d->rd_bcache_cache_allocations_clean, clean);
rrddim_set_by_pointer(d->st_bcache_cache_allocations, d->rd_bcache_cache_allocations_metadata, metadata);
rrddim_set_by_pointer(d->st_bcache_cache_allocations, d->rd_bcache_cache_allocations_unknown, unknown);
rrdset_done(d->st_bcache_cache_allocations);
}
}
static inline int is_major_enabled(int major) {
static int8_t *major_configs = NULL;
static size_t major_size = 0;
if(major < 0) return 1;
size_t wanted_size = (size_t)major + 1;
if(major_size < wanted_size) {
major_configs = reallocz(major_configs, wanted_size * sizeof(int8_t));
size_t i;
for(i = major_size; i < wanted_size ; i++)
major_configs[i] = -1;
major_size = wanted_size;
}
if(major_configs[major] == -1) {
char buffer[CONFIG_MAX_NAME + 1];
snprintfz(buffer, CONFIG_MAX_NAME, "performance metrics for disks with major %d", major);
major_configs[major] = (char)config_get_boolean(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, buffer, 1);
}
return (int)major_configs[major];
}
static inline int get_disk_name_from_path(const char *path, char *result, size_t result_size, unsigned long major, unsigned long minor, char *disk, char *prefix, int depth) {
//info("DEVICE-MAPPER ('%s', %lu:%lu): examining directory '%s' (allowed depth %d).", disk, major, minor, path, depth);
int found = 0, preferred = 0;
char *first_result = mallocz(result_size);
DIR *dir = opendir(path);
if (!dir) {
error("DEVICE-MAPPER ('%s', %lu:%lu): Cannot open directory '%s'.", disk, major, minor, path);
goto failed;
}
struct dirent *de = NULL;
while ((de = readdir(dir))) {
if(de->d_type == DT_DIR) {
if((de->d_name[0] == '.' && de->d_name[1] == '\0') || (de->d_name[0] == '.' && de->d_name[1] == '.' && de->d_name[2] == '\0'))
continue;
if(depth <= 0) {
error("DEVICE-MAPPER ('%s', %lu:%lu): Depth limit reached for path '%s/%s'. Ignoring path.", disk, major, minor, path, de->d_name);
break;
}
else {
char *path_nested = NULL;
char *prefix_nested = NULL;
{
char buffer[FILENAME_MAX + 1];
snprintfz(buffer, FILENAME_MAX, "%s/%s", path, de->d_name);
path_nested = strdupz(buffer);
snprintfz(buffer, FILENAME_MAX, "%s%s%s", (prefix)?prefix:"", (prefix)?"_":"", de->d_name);
prefix_nested = strdupz(buffer);
}
found = get_disk_name_from_path(path_nested, result, result_size, major, minor, disk, prefix_nested, depth - 1);
freez(path_nested);
freez(prefix_nested);
if(found) break;
}
}
else if(de->d_type == DT_LNK || de->d_type == DT_BLK) {
char filename[FILENAME_MAX + 1];
if(de->d_type == DT_LNK) {
snprintfz(filename, FILENAME_MAX, "%s/%s", path, de->d_name);
ssize_t len = readlink(filename, result, result_size - 1);
if(len <= 0) {
error("DEVICE-MAPPER ('%s', %lu:%lu): Cannot read link '%s'.", disk, major, minor, filename);
continue;
}
result[len] = '\0';
if(result[0] != '/')
snprintfz(filename, FILENAME_MAX, "%s/%s", path, result);
else
strncpyz(filename, result, FILENAME_MAX);
}
else {
snprintfz(filename, FILENAME_MAX, "%s/%s", path, de->d_name);
}
struct stat sb;
if(stat(filename, &sb) == -1) {
error("DEVICE-MAPPER ('%s', %lu:%lu): Cannot stat() file '%s'.", disk, major, minor, filename);
continue;
}
if((sb.st_mode & S_IFMT) != S_IFBLK) {
//info("DEVICE-MAPPER ('%s', %lu:%lu): file '%s' is not a block device.", disk, major, minor, filename);
continue;
}
if(major(sb.st_rdev) != major || minor(sb.st_rdev) != minor) {
//info("DEVICE-MAPPER ('%s', %lu:%lu): filename '%s' does not match %lu:%lu.", disk, major, minor, filename, (unsigned long)major(sb.st_rdev), (unsigned long)minor(sb.st_rdev));
continue;
}
//info("DEVICE-MAPPER ('%s', %lu:%lu): filename '%s' matches.", disk, major, minor, filename);
snprintfz(result, result_size - 1, "%s%s%s", (prefix)?prefix:"", (prefix)?"_":"", de->d_name);
if(!found) {
strncpyz(first_result, result, result_size);
found = 1;
}
if(simple_pattern_matches(preferred_ids, result)) {
preferred = 1;
break;
}
}
}
closedir(dir);
failed:
if(!found)
result[0] = '\0';
else if(!preferred)
strncpyz(result, first_result, result_size);
freez(first_result);
return found;
}
static inline char *get_disk_name(unsigned long major, unsigned long minor, char *disk) {
char result[FILENAME_MAX + 1] = "";
if(!path_to_device_mapper || !*path_to_device_mapper || !get_disk_name_from_path(path_to_device_mapper, result, FILENAME_MAX + 1, major, minor, disk, NULL, 0))
if(!path_to_device_label || !*path_to_device_label || !get_disk_name_from_path(path_to_device_label, result, FILENAME_MAX + 1, major, minor, disk, NULL, 0))
if(!path_to_veritas_volume_groups || !*path_to_veritas_volume_groups || !get_disk_name_from_path(path_to_veritas_volume_groups, result, FILENAME_MAX + 1, major, minor, disk, "vx", 2))
if(name_disks_by_id != CONFIG_BOOLEAN_YES || !path_to_device_id || !*path_to_device_id || !get_disk_name_from_path(path_to_device_id, result, FILENAME_MAX + 1, major, minor, disk, NULL, 0))
strncpy(result, disk, FILENAME_MAX);
if(!result[0])
strncpy(result, disk, FILENAME_MAX);
netdata_fix_chart_name(result);
return strdup(result);
}
static void get_disk_config(struct disk *d) {
int def_enable = global_enable_new_disks_detected_at_runtime;
if(def_enable != CONFIG_BOOLEAN_NO && (simple_pattern_matches(excluded_disks, d->device) || simple_pattern_matches(excluded_disks, d->disk)))
def_enable = CONFIG_BOOLEAN_NO;
char var_name[4096 + 1];
snprintfz(var_name, 4096, CONFIG_SECTION_PLUGIN_PROC_DISKSTATS ":%s", d->disk);
def_enable = config_get_boolean_ondemand(var_name, "enable", def_enable);
if(unlikely(def_enable == CONFIG_BOOLEAN_NO)) {
// the user does not want any metrics for this disk
d->do_io = CONFIG_BOOLEAN_NO;
d->do_ops = CONFIG_BOOLEAN_NO;
d->do_mops = CONFIG_BOOLEAN_NO;
d->do_iotime = CONFIG_BOOLEAN_NO;
d->do_qops = CONFIG_BOOLEAN_NO;
d->do_util = CONFIG_BOOLEAN_NO;
d->do_backlog = CONFIG_BOOLEAN_NO;
d->do_bcache = CONFIG_BOOLEAN_NO;
}
else {
// this disk is enabled
// check its direct settings
int def_performance = CONFIG_BOOLEAN_AUTO;
// since this is 'on demand' we can figure the performance settings
// based on the type of disk
if(!d->device_is_bcache) {
switch(d->type) {
default:
case DISK_TYPE_UNKNOWN:
break;
case DISK_TYPE_PHYSICAL:
def_performance = global_enable_performance_for_physical_disks;
break;
case DISK_TYPE_PARTITION:
def_performance = global_enable_performance_for_partitions;
break;
case DISK_TYPE_VIRTUAL:
def_performance = global_enable_performance_for_virtual_disks;
break;
}
}
// check if we have to disable performance for this disk
if(def_performance)
def_performance = is_major_enabled((int)d->major);
// ------------------------------------------------------------
// now we have def_performance and def_space
// to work further
// def_performance
// check the user configuration (this will also show our 'on demand' decision)
def_performance = config_get_boolean_ondemand(var_name, "enable performance metrics", def_performance);
int ddo_io = CONFIG_BOOLEAN_NO,
ddo_ops = CONFIG_BOOLEAN_NO,
ddo_mops = CONFIG_BOOLEAN_NO,
ddo_iotime = CONFIG_BOOLEAN_NO,
ddo_qops = CONFIG_BOOLEAN_NO,
ddo_util = CONFIG_BOOLEAN_NO,
ddo_backlog = CONFIG_BOOLEAN_NO,
ddo_bcache = CONFIG_BOOLEAN_NO;
// we enable individual performance charts only when def_performance is not disabled
if(unlikely(def_performance != CONFIG_BOOLEAN_NO)) {
ddo_io = global_do_io,
ddo_ops = global_do_ops,
ddo_mops = global_do_mops,
ddo_iotime = global_do_iotime,
ddo_qops = global_do_qops,
ddo_util = global_do_util,
ddo_backlog = global_do_backlog,
ddo_bcache = global_do_bcache;
}
d->do_io = config_get_boolean_ondemand(var_name, "bandwidth", ddo_io);
d->do_ops = config_get_boolean_ondemand(var_name, "operations", ddo_ops);
d->do_mops = config_get_boolean_ondemand(var_name, "merged operations", ddo_mops);
d->do_iotime = config_get_boolean_ondemand(var_name, "i/o time", ddo_iotime);
d->do_qops = config_get_boolean_ondemand(var_name, "queued operations", ddo_qops);
d->do_util = config_get_boolean_ondemand(var_name, "utilization percentage", ddo_util);
d->do_backlog = config_get_boolean_ondemand(var_name, "backlog", ddo_backlog);
if(d->device_is_bcache)
d->do_bcache = config_get_boolean_ondemand(var_name, "bcache", ddo_bcache);
else
d->do_bcache = 0;
}
}
static struct disk *get_disk(unsigned long major, unsigned long minor, char *disk) {
static struct mountinfo *disk_mountinfo_root = NULL;
struct disk *d;
// search for it in our RAM list.
// this is sequential, but since we just walk through
// and the number of disks / partitions in a system
// should not be that many, it should be acceptable
for(d = disk_root; d ; d = d->next)
if(unlikely(d->major == major && d->minor == minor))
return d;
// not found
// create a new disk structure
d = (struct disk *)callocz(1, sizeof(struct disk));
d->disk = get_disk_name(major, minor, disk);
d->device = strdupz(disk);
d->major = major;
d->minor = minor;
d->type = DISK_TYPE_UNKNOWN; // Default type. Changed later if not correct.
d->sector_size = 512; // the default, will be changed below
d->next = NULL;
// append it to the list
if(unlikely(!disk_root))
disk_root = d;
else {
struct disk *last;
for(last = disk_root; last->next ;last = last->next);
last->next = d;
}
char buffer[FILENAME_MAX + 1];
// find if it is a physical disk
// by checking if /sys/block/DISK is readable.
snprintfz(buffer, FILENAME_MAX, path_to_sys_block_device, disk);
if(likely(access(buffer, R_OK) == 0)) {
// assign it here, but it will be overwritten if it is not a physical disk
d->type = DISK_TYPE_PHYSICAL;
}
// find if it is a partition
// by checking if /sys/dev/block/MAJOR:MINOR/partition is readable.
snprintfz(buffer, FILENAME_MAX, path_to_sys_dev_block_major_minor_string, major, minor, "partition");
if(likely(access(buffer, R_OK) == 0)) {
d->type = DISK_TYPE_PARTITION;
}
else {
// find if it is a virtual disk
// by checking if /sys/devices/virtual/block/DISK is readable.
snprintfz(buffer, FILENAME_MAX, path_to_sys_devices_virtual_block_device, disk);
if(likely(access(buffer, R_OK) == 0)) {
d->type = DISK_TYPE_VIRTUAL;
}
else {
// find if it is a virtual device
// by checking if /sys/dev/block/MAJOR:MINOR/slaves has entries
snprintfz(buffer, FILENAME_MAX, path_to_sys_dev_block_major_minor_string, major, minor, "slaves/");
DIR *dirp = opendir(buffer);
if (likely(dirp != NULL)) {
struct dirent *dp;
while ((dp = readdir(dirp))) {
// . and .. are also files in empty folders.
if (unlikely(strcmp(dp->d_name, ".") == 0 || strcmp(dp->d_name, "..") == 0)) {
continue;
}
d->type = DISK_TYPE_VIRTUAL;
// Stop the loop after we found one file.
break;
}
if (unlikely(closedir(dirp) == -1))
error("Unable to close dir %s", buffer);
}
}
}
// ------------------------------------------------------------------------
// check if we can find its mount point
// mountinfo_find() can be called with NULL disk_mountinfo_root
struct mountinfo *mi = mountinfo_find(disk_mountinfo_root, d->major, d->minor);
if(unlikely(!mi)) {
// mountinfo_free_all can be called with NULL
mountinfo_free_all(disk_mountinfo_root);
disk_mountinfo_root = mountinfo_read(0);
mi = mountinfo_find(disk_mountinfo_root, d->major, d->minor);
}
if(unlikely(mi))
d->mount_point = strdupz(mi->mount_point);
else
d->mount_point = NULL;
// ------------------------------------------------------------------------
// find the disk sector size
/*
* sector size is always 512 bytes inside the kernel #3481
*
{
char tf[FILENAME_MAX + 1], *t;
strncpyz(tf, d->device, FILENAME_MAX);
// replace all / with !
for(t = tf; *t ;t++)
if(unlikely(*t == '/')) *t = '!';
if(likely(d->type == DISK_TYPE_PARTITION))
snprintfz(buffer, FILENAME_MAX, path_to_get_hw_sector_size_partitions, d->major, d->minor, tf);
else
snprintfz(buffer, FILENAME_MAX, path_to_get_hw_sector_size, tf);
FILE *fpss = fopen(buffer, "r");
if(likely(fpss)) {
char buffer2[1024 + 1];
char *tmp = fgets(buffer2, 1024, fpss);
if(likely(tmp)) {
d->sector_size = str2i(tmp);
if(unlikely(d->sector_size <= 0)) {
error("Invalid sector size %d for device %s in %s. Assuming 512.", d->sector_size, d->device, buffer);
d->sector_size = 512;
}
}
else error("Cannot read data for sector size for device %s from %s. Assuming 512.", d->device, buffer);
fclose(fpss);
}
else error("Cannot read sector size for device %s from %s. Assuming 512.", d->device, buffer);
}
*/
// ------------------------------------------------------------------------
// check if the device is a bcache
struct stat bcache;
snprintfz(buffer, FILENAME_MAX, path_to_sys_block_device_bcache, disk);
if(unlikely(stat(buffer, &bcache) == 0 && (bcache.st_mode & S_IFMT) == S_IFDIR)) {
// we have the 'bcache' directory
d->device_is_bcache = 1;
char buffer2[FILENAME_MAX + 1];
snprintfz(buffer2, FILENAME_MAX, "%s/cache/congested", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_cache_congested = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/readahead", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_stats_total_cache_readaheads = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/cache/cache0/priority_stats", buffer); // only one cache is supported by bcache
if(access(buffer2, R_OK) == 0)
d->bcache_filename_priority_stats = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/cache/internal/cache_read_races", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_cache_read_races = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/cache/cache0/io_errors", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_cache_io_errors = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/dirty_data", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_dirty_data = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/writeback_rate", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_writeback_rate = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/cache/cache_available_percent", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_cache_available_percent = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/stats_total/cache_hits", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_stats_total_cache_hits = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/stats_five_minute/cache_hit_ratio", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_stats_five_minute_cache_hit_ratio = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/stats_hour/cache_hit_ratio", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_stats_hour_cache_hit_ratio = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/stats_day/cache_hit_ratio", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_stats_day_cache_hit_ratio = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/stats_total/cache_hit_ratio", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_stats_total_cache_hit_ratio = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/stats_total/cache_misses", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_stats_total_cache_misses = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/stats_total/cache_bypass_hits", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_stats_total_cache_bypass_hits = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/stats_total/cache_bypass_misses", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_stats_total_cache_bypass_misses = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
snprintfz(buffer2, FILENAME_MAX, "%s/stats_total/cache_miss_collisions", buffer);
if(access(buffer2, R_OK) == 0)
d->bcache_filename_stats_total_cache_miss_collisions = strdupz(buffer2);
else
error("bcache file '%s' cannot be read.", buffer2);
}
get_disk_config(d);
return d;
}
int do_proc_diskstats(int update_every, usec_t dt) {
static procfile *ff = NULL;
if(unlikely(!globals_initialized)) {
globals_initialized = 1;
global_enable_new_disks_detected_at_runtime = config_get_boolean(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "enable new disks detected at runtime", global_enable_new_disks_detected_at_runtime);
global_enable_performance_for_physical_disks = config_get_boolean_ondemand(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "performance metrics for physical disks", global_enable_performance_for_physical_disks);
global_enable_performance_for_virtual_disks = config_get_boolean_ondemand(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "performance metrics for virtual disks", global_enable_performance_for_virtual_disks);
global_enable_performance_for_partitions = config_get_boolean_ondemand(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "performance metrics for partitions", global_enable_performance_for_partitions);
global_do_io = config_get_boolean_ondemand(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "bandwidth for all disks", global_do_io);
global_do_ops = config_get_boolean_ondemand(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "operations for all disks", global_do_ops);
global_do_mops = config_get_boolean_ondemand(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "merged operations for all disks", global_do_mops);
global_do_iotime = config_get_boolean_ondemand(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "i/o time for all disks", global_do_iotime);
global_do_qops = config_get_boolean_ondemand(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "queued operations for all disks", global_do_qops);
global_do_util = config_get_boolean_ondemand(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "utilization percentage for all disks", global_do_util);
global_do_backlog = config_get_boolean_ondemand(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "backlog for all disks", global_do_backlog);
global_do_bcache = config_get_boolean_ondemand(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "bcache for all disks", global_do_bcache);
global_bcache_priority_stats_update_every = (int)config_get_number(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "bcache priority stats update every", global_bcache_priority_stats_update_every);
global_cleanup_removed_disks = config_get_boolean(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "remove charts of removed disks" , global_cleanup_removed_disks);
char buffer[FILENAME_MAX + 1];
snprintfz(buffer, FILENAME_MAX, "%s%s", netdata_configured_host_prefix, "/sys/block/%s");
path_to_sys_block_device = config_get(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "path to get block device", buffer);
snprintfz(buffer, FILENAME_MAX, "%s%s", netdata_configured_host_prefix, "/sys/block/%s/bcache");
path_to_sys_block_device_bcache = config_get(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "path to get block device bcache", buffer);
snprintfz(buffer, FILENAME_MAX, "%s%s", netdata_configured_host_prefix, "/sys/devices/virtual/block/%s");
path_to_sys_devices_virtual_block_device = config_get(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "path to get virtual block device", buffer);
snprintfz(buffer, FILENAME_MAX, "%s%s", netdata_configured_host_prefix, "/sys/dev/block/%lu:%lu/%s");
path_to_sys_dev_block_major_minor_string = config_get(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "path to get block device infos", buffer);
//snprintfz(buffer, FILENAME_MAX, "%s%s", netdata_configured_host_prefix, "/sys/block/%s/queue/hw_sector_size");
//path_to_get_hw_sector_size = config_get(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "path to get h/w sector size", buffer);
//snprintfz(buffer, FILENAME_MAX, "%s%s", netdata_configured_host_prefix, "/sys/dev/block/%lu:%lu/subsystem/%s/../queue/hw_sector_size");
//path_to_get_hw_sector_size_partitions = config_get(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "path to get h/w sector size for partitions", buffer);
snprintfz(buffer, FILENAME_MAX, "%s/dev/mapper", netdata_configured_host_prefix);
path_to_device_mapper = config_get(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "path to device mapper", buffer);
snprintfz(buffer, FILENAME_MAX, "%s/dev/disk/by-label", netdata_configured_host_prefix);
path_to_device_label = config_get(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "path to /dev/disk/by-label", buffer);
snprintfz(buffer, FILENAME_MAX, "%s/dev/disk/by-id", netdata_configured_host_prefix);
path_to_device_id = config_get(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "path to /dev/disk/by-id", buffer);
snprintfz(buffer, FILENAME_MAX, "%s/dev/vx/dsk", netdata_configured_host_prefix);
path_to_veritas_volume_groups = config_get(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "path to /dev/vx/dsk", buffer);
name_disks_by_id = config_get_boolean(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "name disks by id", name_disks_by_id);
preferred_ids = simple_pattern_create(
config_get(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "preferred disk ids", DEFAULT_PREFERRED_IDS)
, NULL
, SIMPLE_PATTERN_EXACT
);
excluded_disks = simple_pattern_create(
config_get(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "exclude disks", DEFAULT_EXCLUDED_DISKS)
, NULL
, SIMPLE_PATTERN_EXACT
);
}
// --------------------------------------------------------------------------
if(unlikely(!ff)) {
char filename[FILENAME_MAX + 1];
snprintfz(filename, FILENAME_MAX, "%s%s", netdata_configured_host_prefix, "/proc/diskstats");
ff = procfile_open(config_get(CONFIG_SECTION_PLUGIN_PROC_DISKSTATS, "filename to monitor", filename), " \t", PROCFILE_FLAG_DEFAULT);
}
if(unlikely(!ff)) return 0;
ff = procfile_readall(ff);
if(unlikely(!ff)) return 0; // we return 0, so that we will retry to open it next time
size_t lines = procfile_lines(ff), l;
collected_number system_read_kb = 0, system_write_kb = 0;
for(l = 0; l < lines ;l++) {
// --------------------------------------------------------------------------
// Read parameters
char *disk;
unsigned long major = 0, minor = 0;
collected_number reads = 0, mreads = 0, readsectors = 0, readms = 0,
writes = 0, mwrites = 0, writesectors = 0, writems = 0,
queued_ios = 0, busy_ms = 0, backlog_ms = 0;
collected_number last_reads = 0, last_readsectors = 0, last_readms = 0,
last_writes = 0, last_writesectors = 0, last_writems = 0,
last_busy_ms = 0;
size_t words = procfile_linewords(ff, l);
if(unlikely(words < 14)) continue;
major = str2ul(procfile_lineword(ff, l, 0));
minor = str2ul(procfile_lineword(ff, l, 1));
disk = procfile_lineword(ff, l, 2);
// # of reads completed # of writes completed
// This is the total number of reads or writes completed successfully.
reads = str2ull(procfile_lineword(ff, l, 3)); // rd_ios
writes = str2ull(procfile_lineword(ff, l, 7)); // wr_ios
// # of reads merged # of writes merged
// Reads and writes which are adjacent to each other may be merged for
// efficiency. Thus two 4K reads may become one 8K read before it is
// ultimately handed to the disk, and so it will be counted (and queued)
mreads = str2ull(procfile_lineword(ff, l, 4)); // rd_merges_or_rd_sec
mwrites = str2ull(procfile_lineword(ff, l, 8)); // wr_merges
// # of sectors read # of sectors written
// This is the total number of sectors read or written successfully.
readsectors = str2ull(procfile_lineword(ff, l, 5)); // rd_sec_or_wr_ios
writesectors = str2ull(procfile_lineword(ff, l, 9)); // wr_sec
// # of milliseconds spent reading # of milliseconds spent writing
// This is the total number of milliseconds spent by all reads or writes (as
// measured from __make_request() to end_that_request_last()).
readms = str2ull(procfile_lineword(ff, l, 6)); // rd_ticks_or_wr_sec
writems = str2ull(procfile_lineword(ff, l, 10)); // wr_ticks
// # of I/Os currently in progress
// The only field that should go to zero. Incremented as requests are
// given to appropriate struct request_queue and decremented as they finish.
queued_ios = str2ull(procfile_lineword(ff, l, 11)); // ios_pgr
// # of milliseconds spent doing I/Os
// This field increases so long as field queued_ios is nonzero.
busy_ms = str2ull(procfile_lineword(ff, l, 12)); // tot_ticks
// weighted # of milliseconds spent doing I/Os
// This field is incremented at each I/O start, I/O completion, I/O
// merge, or read of these stats by the number of I/Os in progress
// (field queued_ios) times the number of milliseconds spent doing I/O since the
// last update of this field. This can provide an easy measure of both
// I/O completion time and the backlog that may be accumulating.
backlog_ms = str2ull(procfile_lineword(ff, l, 13)); // rq_ticks
// --------------------------------------------------------------------------
// remove slashes from disk names
char *s;
for(s = disk; *s ;s++)
if(*s == '/') *s = '_';
// --------------------------------------------------------------------------
// get a disk structure for the disk
struct disk *d = get_disk(major, minor, disk);
d->updated = 1;
// --------------------------------------------------------------------------
// count the global system disk I/O of physical disks
if(unlikely(d->type == DISK_TYPE_PHYSICAL)) {
system_read_kb += readsectors * d->sector_size / 1024;
system_write_kb += writesectors * d->sector_size / 1024;
}
// --------------------------------------------------------------------------
// Set its family based on mount point
char *family = d->mount_point;
if(!family) family = d->disk;
// --------------------------------------------------------------------------
// Do performance metrics
if(d->do_io == CONFIG_BOOLEAN_YES || (d->do_io == CONFIG_BOOLEAN_AUTO &&
(readsectors || writesectors ||
netdata_zero_metrics_enabled == CONFIG_BOOLEAN_YES))) {
d->do_io = CONFIG_BOOLEAN_YES;
if(unlikely(!d->st_io)) {
d->st_io = rrdset_create_localhost(
RRD_TYPE_DISK
, d->device
, d->disk
, family
, "disk.io"
, "Disk I/O Bandwidth"
, "KiB/s"
, PLUGIN_PROC_NAME
, PLUGIN_PROC_MODULE_DISKSTATS_NAME
, NETDATA_CHART_PRIO_DISK_IO
, update_every
, RRDSET_TYPE_AREA
);
d->rd_io_reads = rrddim_add(d->st_io, "reads", NULL, d->sector_size, 1024, RRD_ALGORITHM_INCREMENTAL);
d->rd_io_writes = rrddim_add(d->st_io, "writes", NULL, d->sector_size * -1, 1024, RRD_ALGORITHM_INCREMENTAL);
}
else rrdset_next(d->st_io);