forked from axboe/fio
-
Notifications
You must be signed in to change notification settings - Fork 0
/
eta.c
630 lines (536 loc) · 14.5 KB
/
eta.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
/*
* Status and ETA code
*/
#include <unistd.h>
#include <fcntl.h>
#include <string.h>
#include "fio.h"
#include "lib/pow2.h"
static char __run_str[REAL_MAX_JOBS + 1];
static char run_str[__THREAD_RUNSTR_SZ(REAL_MAX_JOBS)];
static void update_condensed_str(char *rstr, char *run_str_condensed)
{
if (*rstr) {
while (*rstr) {
int nr = 1;
*run_str_condensed++ = *rstr++;
while (*(rstr - 1) == *rstr) {
rstr++;
nr++;
}
run_str_condensed += sprintf(run_str_condensed, "(%u),", nr);
}
run_str_condensed--;
}
*run_str_condensed = '\0';
}
/*
* Sets the status of the 'td' in the printed status map.
*/
static void check_str_update(struct thread_data *td)
{
char c = __run_str[td->thread_number - 1];
switch (td->runstate) {
case TD_REAPED:
if (td->error)
c = 'X';
else if (td->sig)
c = 'K';
else
c = '_';
break;
case TD_EXITED:
c = 'E';
break;
case TD_RAMP:
c = '/';
break;
case TD_RUNNING:
if (td_rw(td)) {
if (td_random(td)) {
if (td->o.rwmix[DDIR_READ] == 100)
c = 'r';
else if (td->o.rwmix[DDIR_WRITE] == 100)
c = 'w';
else
c = 'm';
} else {
if (td->o.rwmix[DDIR_READ] == 100)
c = 'R';
else if (td->o.rwmix[DDIR_WRITE] == 100)
c = 'W';
else
c = 'M';
}
} else if (td_read(td)) {
if (td_random(td))
c = 'r';
else
c = 'R';
} else if (td_write(td)) {
if (td_random(td))
c = 'w';
else
c = 'W';
} else {
if (td_random(td))
c = 'd';
else
c = 'D';
}
break;
case TD_PRE_READING:
c = 'p';
break;
case TD_VERIFYING:
c = 'V';
break;
case TD_FSYNCING:
c = 'F';
break;
case TD_FINISHING:
c = 'f';
break;
case TD_CREATED:
c = 'C';
break;
case TD_INITIALIZED:
case TD_SETTING_UP:
c = 'I';
break;
case TD_NOT_CREATED:
c = 'P';
break;
default:
log_err("state %d\n", td->runstate);
}
__run_str[td->thread_number - 1] = c;
update_condensed_str(__run_str, run_str);
}
/*
* Convert seconds to a printable string.
*/
void eta_to_str(char *str, unsigned long eta_sec)
{
unsigned int d, h, m, s;
int disp_hour = 0;
if (eta_sec == -1) {
sprintf(str, "--");
return;
}
s = eta_sec % 60;
eta_sec /= 60;
m = eta_sec % 60;
eta_sec /= 60;
h = eta_sec % 24;
eta_sec /= 24;
d = eta_sec;
if (d) {
disp_hour = 1;
str += sprintf(str, "%02ud:", d);
}
if (h || disp_hour)
str += sprintf(str, "%02uh:", h);
str += sprintf(str, "%02um:", m);
str += sprintf(str, "%02us", s);
}
/*
* Best effort calculation of the estimated pending runtime of a job.
*/
static unsigned long thread_eta(struct thread_data *td)
{
unsigned long long bytes_total, bytes_done;
unsigned long eta_sec = 0;
unsigned long elapsed;
uint64_t timeout;
elapsed = (mtime_since_now(&td->epoch) + 999) / 1000;
timeout = td->o.timeout / 1000000UL;
bytes_total = td->total_io_size;
if (td->flags & TD_F_NO_PROGRESS)
return -1;
if (td->o.fill_device && td->o.size == -1ULL) {
if (!td->fill_device_size || td->fill_device_size == -1ULL)
return 0;
bytes_total = td->fill_device_size;
}
if (td->o.zone_size && td->o.zone_skip && bytes_total) {
unsigned int nr_zones;
uint64_t zone_bytes;
zone_bytes = bytes_total + td->o.zone_size + td->o.zone_skip;
nr_zones = (zone_bytes - 1) / (td->o.zone_size + td->o.zone_skip);
bytes_total -= nr_zones * td->o.zone_skip;
}
/*
* if writing and verifying afterwards, bytes_total will be twice the
* size. In a mixed workload, verify phase will be the size of the
* first stage writes.
*/
if (td->o.do_verify && td->o.verify && td_write(td)) {
if (td_rw(td)) {
unsigned int perc = 50;
if (td->o.rwmix[DDIR_WRITE])
perc = td->o.rwmix[DDIR_WRITE];
bytes_total += (bytes_total * perc) / 100;
} else
bytes_total <<= 1;
}
if (td->runstate == TD_RUNNING || td->runstate == TD_VERIFYING) {
double perc, perc_t;
bytes_done = ddir_rw_sum(td->io_bytes);
if (bytes_total) {
perc = (double) bytes_done / (double) bytes_total;
if (perc > 1.0)
perc = 1.0;
} else
perc = 0.0;
if (td->o.time_based) {
if (timeout) {
perc_t = (double) elapsed / (double) timeout;
if (perc_t < perc)
perc = perc_t;
} else {
/*
* Will never hit, we can't have time_based
* without a timeout set.
*/
perc = 0.0;
}
}
eta_sec = (unsigned long) (elapsed * (1.0 / perc)) - elapsed;
if (td->o.timeout &&
eta_sec > (timeout + done_secs - elapsed))
eta_sec = timeout + done_secs - elapsed;
} else if (td->runstate == TD_NOT_CREATED || td->runstate == TD_CREATED
|| td->runstate == TD_INITIALIZED
|| td->runstate == TD_SETTING_UP
|| td->runstate == TD_RAMP
|| td->runstate == TD_PRE_READING) {
int t_eta = 0, r_eta = 0;
unsigned long long rate_bytes;
/*
* We can only guess - assume it'll run the full timeout
* if given, otherwise assume it'll run at the specified rate.
*/
if (td->o.timeout) {
uint64_t __timeout = td->o.timeout;
uint64_t start_delay = td->o.start_delay;
uint64_t ramp_time = td->o.ramp_time;
t_eta = __timeout + start_delay + ramp_time;
t_eta /= 1000000ULL;
if ((td->runstate == TD_RAMP) && in_ramp_time(td)) {
unsigned long ramp_left;
ramp_left = mtime_since_now(&td->epoch);
ramp_left = (ramp_left + 999) / 1000;
if (ramp_left <= t_eta)
t_eta -= ramp_left;
}
}
rate_bytes = ddir_rw_sum(td->o.rate);
if (rate_bytes) {
r_eta = (bytes_total / 1024) / rate_bytes;
r_eta += (td->o.start_delay / 1000000ULL);
}
if (r_eta && t_eta)
eta_sec = min(r_eta, t_eta);
else if (r_eta)
eta_sec = r_eta;
else if (t_eta)
eta_sec = t_eta;
else
eta_sec = 0;
} else {
/*
* thread is already done or waiting for fsync
*/
eta_sec = 0;
}
return eta_sec;
}
static void calc_rate(int unified_rw_rep, unsigned long mtime,
unsigned long long *io_bytes,
unsigned long long *prev_io_bytes, unsigned int *rate)
{
int i;
for (i = 0; i < DDIR_RWDIR_CNT; i++) {
unsigned long long diff, this_rate;
diff = io_bytes[i] - prev_io_bytes[i];
if (mtime)
this_rate = ((1000 * diff) / mtime) / 1024;
else
this_rate = 0;
if (unified_rw_rep) {
rate[i] = 0;
rate[0] += this_rate;
} else
rate[i] = this_rate;
prev_io_bytes[i] = io_bytes[i];
}
}
static void calc_iops(int unified_rw_rep, unsigned long mtime,
unsigned long long *io_iops,
unsigned long long *prev_io_iops, unsigned int *iops)
{
int i;
for (i = 0; i < DDIR_RWDIR_CNT; i++) {
unsigned long long diff, this_iops;
diff = io_iops[i] - prev_io_iops[i];
if (mtime)
this_iops = (diff * 1000) / mtime;
else
this_iops = 0;
if (unified_rw_rep) {
iops[i] = 0;
iops[0] += this_iops;
} else
iops[i] = this_iops;
prev_io_iops[i] = io_iops[i];
}
}
/*
* Print status of the jobs we know about. This includes rate estimates,
* ETA, thread state, etc.
*/
int calc_thread_status(struct jobs_eta *je, int force)
{
struct thread_data *td;
int i, unified_rw_rep;
unsigned long rate_time, disp_time, bw_avg_time, *eta_secs;
unsigned long long io_bytes[DDIR_RWDIR_CNT];
unsigned long long io_iops[DDIR_RWDIR_CNT];
struct timeval now;
static unsigned long long rate_io_bytes[DDIR_RWDIR_CNT];
static unsigned long long disp_io_bytes[DDIR_RWDIR_CNT];
static unsigned long long disp_io_iops[DDIR_RWDIR_CNT];
static struct timeval rate_prev_time, disp_prev_time;
if (!force) {
if (!(output_format & FIO_OUTPUT_NORMAL) &&
f_out == stdout)
return 0;
if (temp_stall_ts || eta_print == FIO_ETA_NEVER)
return 0;
if (!isatty(STDOUT_FILENO) && (eta_print != FIO_ETA_ALWAYS))
return 0;
}
if (!ddir_rw_sum(rate_io_bytes))
fill_start_time(&rate_prev_time);
if (!ddir_rw_sum(disp_io_bytes))
fill_start_time(&disp_prev_time);
eta_secs = malloc(thread_number * sizeof(unsigned long));
memset(eta_secs, 0, thread_number * sizeof(unsigned long));
je->elapsed_sec = (mtime_since_genesis() + 999) / 1000;
io_bytes[DDIR_READ] = io_bytes[DDIR_WRITE] = io_bytes[DDIR_TRIM] = 0;
io_iops[DDIR_READ] = io_iops[DDIR_WRITE] = io_iops[DDIR_TRIM] = 0;
bw_avg_time = ULONG_MAX;
unified_rw_rep = 0;
for_each_td(td, i) {
unified_rw_rep += td->o.unified_rw_rep;
if (is_power_of_2(td->o.kb_base))
je->is_pow2 = 1;
je->unit_base = td->o.unit_base;
if (td->o.bw_avg_time < bw_avg_time)
bw_avg_time = td->o.bw_avg_time;
if (td->runstate == TD_RUNNING || td->runstate == TD_VERIFYING
|| td->runstate == TD_FSYNCING
|| td->runstate == TD_PRE_READING
|| td->runstate == TD_FINISHING) {
je->nr_running++;
if (td_read(td)) {
je->t_rate[0] += td->o.rate[DDIR_READ];
je->t_iops[0] += td->o.rate_iops[DDIR_READ];
je->m_rate[0] += td->o.ratemin[DDIR_READ];
je->m_iops[0] += td->o.rate_iops_min[DDIR_READ];
}
if (td_write(td)) {
je->t_rate[1] += td->o.rate[DDIR_WRITE];
je->t_iops[1] += td->o.rate_iops[DDIR_WRITE];
je->m_rate[1] += td->o.ratemin[DDIR_WRITE];
je->m_iops[1] += td->o.rate_iops_min[DDIR_WRITE];
}
if (td_trim(td)) {
je->t_rate[2] += td->o.rate[DDIR_TRIM];
je->t_iops[2] += td->o.rate_iops[DDIR_TRIM];
je->m_rate[2] += td->o.ratemin[DDIR_TRIM];
je->m_iops[2] += td->o.rate_iops_min[DDIR_TRIM];
}
je->files_open += td->nr_open_files;
} else if (td->runstate == TD_RAMP) {
je->nr_running++;
je->nr_ramp++;
} else if (td->runstate == TD_SETTING_UP)
je->nr_setting_up++;
else if (td->runstate < TD_RUNNING)
je->nr_pending++;
if (je->elapsed_sec >= 3)
eta_secs[i] = thread_eta(td);
else
eta_secs[i] = INT_MAX;
check_str_update(td);
if (td->runstate > TD_SETTING_UP) {
int ddir;
for (ddir = DDIR_READ; ddir < DDIR_RWDIR_CNT; ddir++) {
if (unified_rw_rep) {
io_bytes[0] += td->io_bytes[ddir];
io_iops[0] += td->io_blocks[ddir];
} else {
io_bytes[ddir] += td->io_bytes[ddir];
io_iops[ddir] += td->io_blocks[ddir];
}
}
}
}
if (exitall_on_terminate) {
je->eta_sec = INT_MAX;
for_each_td(td, i) {
if (eta_secs[i] < je->eta_sec)
je->eta_sec = eta_secs[i];
}
} else {
unsigned long eta_stone = 0;
je->eta_sec = 0;
for_each_td(td, i) {
if ((td->runstate == TD_NOT_CREATED) && td->o.stonewall)
eta_stone += eta_secs[i];
else {
if (eta_secs[i] > je->eta_sec)
je->eta_sec = eta_secs[i];
}
}
je->eta_sec += eta_stone;
}
free(eta_secs);
fio_gettime(&now, NULL);
rate_time = mtime_since(&rate_prev_time, &now);
if (write_bw_log && rate_time > bw_avg_time && !in_ramp_time(td)) {
calc_rate(unified_rw_rep, rate_time, io_bytes, rate_io_bytes,
je->rate);
memcpy(&rate_prev_time, &now, sizeof(now));
add_agg_sample(je->rate[DDIR_READ], DDIR_READ, 0);
add_agg_sample(je->rate[DDIR_WRITE], DDIR_WRITE, 0);
add_agg_sample(je->rate[DDIR_TRIM], DDIR_TRIM, 0);
}
disp_time = mtime_since(&disp_prev_time, &now);
/*
* Allow a little slack, the target is to print it every 1000 msecs
*/
if (!force && disp_time < 900)
return 0;
calc_rate(unified_rw_rep, disp_time, io_bytes, disp_io_bytes, je->rate);
calc_iops(unified_rw_rep, disp_time, io_iops, disp_io_iops, je->iops);
memcpy(&disp_prev_time, &now, sizeof(now));
if (!force && !je->nr_running && !je->nr_pending)
return 0;
je->nr_threads = thread_number;
update_condensed_str(__run_str, run_str);
memcpy(je->run_str, run_str, strlen(run_str));
return 1;
}
void display_thread_status(struct jobs_eta *je)
{
static struct timeval disp_eta_new_line;
static int eta_new_line_init, eta_new_line_pending;
static int linelen_last;
static int eta_good;
char output[REAL_MAX_JOBS + 512], *p = output;
char eta_str[128];
double perc = 0.0;
if (je->eta_sec != INT_MAX && je->elapsed_sec) {
perc = (double) je->elapsed_sec / (double) (je->elapsed_sec + je->eta_sec);
eta_to_str(eta_str, je->eta_sec);
}
if (eta_new_line_pending) {
eta_new_line_pending = 0;
p += sprintf(p, "\n");
}
p += sprintf(p, "Jobs: %d (f=%d)", je->nr_running, je->files_open);
if (je->m_rate[0] || je->m_rate[1] || je->t_rate[0] || je->t_rate[1]) {
char *tr, *mr;
mr = num2str(je->m_rate[0] + je->m_rate[1], 4, 0, je->is_pow2, 8);
tr = num2str(je->t_rate[0] + je->t_rate[1], 4, 0, je->is_pow2, 8);
p += sprintf(p, ", CR=%s/%s KB/s", tr, mr);
free(tr);
free(mr);
} else if (je->m_iops[0] || je->m_iops[1] || je->t_iops[0] || je->t_iops[1]) {
p += sprintf(p, ", CR=%d/%d IOPS",
je->t_iops[0] + je->t_iops[1],
je->m_iops[0] + je->m_iops[1]);
}
if (je->eta_sec != INT_MAX && je->nr_running) {
char perc_str[32];
char *iops_str[DDIR_RWDIR_CNT];
char *rate_str[DDIR_RWDIR_CNT];
size_t left;
int l;
int ddir;
if ((!je->eta_sec && !eta_good) || je->nr_ramp == je->nr_running ||
je->eta_sec == -1)
strcpy(perc_str, "-.-% done");
else {
double mult = 100.0;
if (je->nr_setting_up && je->nr_running)
mult *= (1.0 - (double) je->nr_setting_up / (double) je->nr_running);
eta_good = 1;
perc *= mult;
sprintf(perc_str, "%3.1f%% done", perc);
}
for (ddir = DDIR_READ; ddir < DDIR_RWDIR_CNT; ddir++) {
rate_str[ddir] = num2str(je->rate[ddir], 5,
1024, je->is_pow2, je->unit_base);
iops_str[ddir] = num2str(je->iops[ddir], 4, 1, 0, 0);
}
left = sizeof(output) - (p - output) - 1;
l = snprintf(p, left, ": [%s] [%s] [%s/%s/%s /s] [%s/%s/%s iops] [eta %s]",
je->run_str, perc_str, rate_str[DDIR_READ],
rate_str[DDIR_WRITE], rate_str[DDIR_TRIM],
iops_str[DDIR_READ], iops_str[DDIR_WRITE],
iops_str[DDIR_TRIM], eta_str);
p += l;
if (l >= 0 && l < linelen_last)
p += sprintf(p, "%*s", linelen_last - l, "");
linelen_last = l;
for (ddir = DDIR_READ; ddir < DDIR_RWDIR_CNT; ddir++) {
free(rate_str[ddir]);
free(iops_str[ddir]);
}
}
p += sprintf(p, "\r");
printf("%s", output);
if (!eta_new_line_init) {
fio_gettime(&disp_eta_new_line, NULL);
eta_new_line_init = 1;
} else if (eta_new_line && mtime_since_now(&disp_eta_new_line) > eta_new_line) {
fio_gettime(&disp_eta_new_line, NULL);
eta_new_line_pending = 1;
}
fflush(stdout);
}
struct jobs_eta *get_jobs_eta(bool force, size_t *size)
{
struct jobs_eta *je;
if (!thread_number)
return NULL;
*size = sizeof(*je) + THREAD_RUNSTR_SZ + 8;
je = malloc(*size);
if (!je)
return NULL;
memset(je, 0, *size);
if (!calc_thread_status(je, force)) {
free(je);
return NULL;
}
*size = sizeof(*je) + strlen((char *) je->run_str) + 1;
return je;
}
void print_thread_status(void)
{
struct jobs_eta *je;
size_t size;
je = get_jobs_eta(false, &size);
if (je)
display_thread_status(je);
free(je);
}
void print_status_init(int thr_number)
{
__run_str[thr_number] = 'P';
update_condensed_str(__run_str, run_str);
}