Newer
Older
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
/*
* cputool.c - CPU & load managmenet tool
* Copyright (C) 2012-2013, AllWorldIT
* Copyright (C) 2012, Nigel Kukard <nkukard@lbsd.net>
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#include "config.h"
#include "cputool.h"
#include <dirent.h>
#include <errno.h>
#include <fcntl.h>
#include <getopt.h>
#include <signal.h>
#include <stdarg.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/resource.h>
#include <sys/stat.h>
#include <sys/time.h>
#include <sys/types.h>
#include <sys/wait.h>
#include <time.h>
#include <unistd.h>
/* How accurate is the clock, per second? */
#define CLOCK_PRECISION 1000000
/* How many times do we plan to sleep per second? */
#define DEFAULT_SLEEP CLOCK_PRECISION / 10
/* Be verbose? */
int verbose = 0;
/* These two child variable are used for the signalling function */
int child_pgid = 0;
int child_external = 0;
/* Use with the SIGUSR1 and SIGUSR2 to stop/resume suspending/unsuspending */
int canStopProcesses = 1;
/* Continue loopin flag */
int loop = 1;
/* Counters */
unsigned long statsSlowdowns = 0;
unsigned long statsTimeDelayed = 0;
/* Print out our usage */
static void printUsage(char **argv) {
fprintf(stderr,"Usage: %s [options] [command]\n",argv[0]);
fprintf(stderr,"\n");
fprintf(stderr,"Options:\n");
fprintf(stderr," -p, --pid=<PID> Manage the CPU usage of a specific PID\n");
fprintf(stderr," -c, --cpu-limit=<PCNT> Percentage of CPU to limit process to\n");
fprintf(stderr," -l, --load-limit=<LOAD> Load to limit process to. Decimals allowed\n");
fprintf(stderr," -v, --verbose Be verbose in what we do\n");
fprintf(stderr," -h, --help Display this page\n\n");
fprintf(stderr,"\n");
}
/* Log messsage */
static void logmsg(const char* format, ...)
{
/* Grab time */
time_t t = time(NULL);
struct tm tm = *localtime(&t);
va_list argptr;
/* Print out some fancy info */
fprintf(stderr,"%d-%02d-%02d %02d:%02d:%02d - ", tm.tm_year + 1900, tm.tm_mon + 1, tm.tm_mday, tm.tm_hour, tm.tm_min, tm.tm_sec);
/* And the rest of the mssage */
va_start(argptr, format);
vfprintf(stderr, format, argptr);
va_end(argptr);
}
/* Function to grab the load average */
static float getload() {
float loadavgn = -1.00;
#ifdef __FreeBSD__
long total;
struct loadavg sysload;
size_t size = sizeof(sysload);
static int mib[] = {CTL_VM, VM_LOADAVG};
if (sysctl(mib, 2, &sysload, &size, NULL, 0) < 0) {
(void) fprintf(stderr, "top: sysctl failed: %s\n", strerror(errno));
bzero(&total, sizeof(total));
}
loadavgn = ((double) sysload.ldavg[0]) / sysload.fscale;
#else
char buf[1024];
char *pos;
int fd;
int len;
if ((fd = open("/proc/loadavg", O_RDONLY) > 0)) {
len = read(fd, buf, sizeof(buf) - 1);
close(fd);
if (len > -1) {
buf[len] = '\0';
loadavgn = strtod(buf, &pos);
}
}
#endif
return loadavgn;
}
/* Grab the stat info for a process */
static int getProcessStat(pid_t pid, struct cputool_stat *pstat)
{
char statfile[32];
FILE *fd;
int i;
/* Create filename */
sprintf(statfile, "/proc/%d/stat", pid);
/* Open stat file */
if (!(fd = fopen(statfile,"r"))) {
// logmsg("ERROR: Failed to open '%s': %s\n",statfile,strerror(errno));
return -1;
}
/* Scan in stat */
i = fscanf(fd,CPUTOOL_STAT_FORMAT,
&pstat->pid,pstat->comm,&pstat->state,&pstat->ppid,&pstat->pgrp,
&pstat->session,&pstat->tty_nr,&pstat->tpgid,
&pstat->flags,
&pstat->minflt,&pstat->cminflt,&pstat->majflt,&pstat->cmajflt,
&pstat->utime,&pstat->stime,&pstat->cutime,&pstat->cstime,
&pstat->priority, &pstat->nice,
&pstat->num_threads,
&pstat->itrealvalue,
&pstat->starttime,
&pstat->vsize,&pstat->rss,&pstat->rlim,
&pstat->startcode,&pstat->endcode,&pstat->startstack,
&pstat->kstkesp,&pstat->kstkeip,
&pstat->signal,&pstat->blocked,
&pstat->sigignore,&pstat->sigcatch,
&pstat->wchan,
&pstat->nswap,&pstat->cnswap,
&pstat->exit_signal,
&pstat->processor,&pstat->rt_priority,
&pstat->policy,
&pstat->delayacct_blkio_ticks
);
fclose(fd);
/* Check result */
if (i < 42)
return -1;
return 0;
}
/* Grab an entire process groups CPU time */
static unsigned long getProcessGroupCPUTime(pid_t pgrp)
{
/* Handles & structures */
DIR *proc;
struct dirent *entry = NULL;
struct cputool_stat pstat;
/* Combined total of CPU time consumed */
unsigned long cpuTime = 0;
/* Open /proc */
if ((proc = opendir("/proc")) == NULL) {
logmsg("ERROR: Failed to opendir() on '/proc': %s\n",strerror(errno));
return -1;
}
/* Loop reading in directory entries */
while ((entry = readdir(proc))) {
if (strtok(entry->d_name,"0123456789"))
continue;
/* Grab process stat for this PID */
if (!getProcessStat(atoi(entry->d_name),&pstat)) {
/* If its in our group, then add up the CPU time */
if (pstat.pgrp == pgrp) {
cpuTime += pstat.utime + pstat.stime;
}
}
}
/* Close off our handle in /proc */
closedir(proc);
return cpuTime;
}
/* Return time difference between two timevals in ms */
static inline unsigned long timediff_us(const struct timeval *tv1,const struct timeval *tv2)
{
/* Calculate the total time difference by adding up secs + usecs */
return (tv1->tv_sec - tv2->tv_sec) * 1000000 + (tv1->tv_usec - tv2->tv_usec);
}
/* Signal handling */
static void sigusr1 () {
signal(SIGUSR1, sigusr1);
canStopProcesses = 0;
}
static void sigusr2 () {
signal(SIGUSR2, sigusr2);
canStopProcesses = 1;
}
/* And the handler itself */
static void signal_handler(int signum) {
/* If we have a child we can handle it here */
if(child_pgid > 0) {
/* Resume so the child can handle the signal */
killpg(child_pgid, SIGCONT);
/* Only kill if its ours */
if (!child_external) {
killpg(child_pgid, signum);
waitpid(-1, NULL, 0);
}
}
/* We should not continue looping */
loop = 0;
}
/* Main program */
int main (int argc, char *argv[]) {
/* If this variable is set, we spawned a child */
int haveChild = 0;
/* PID of process we're working on */
pid_t pid = 0;
/* Process GROUP */
pid_t pgid = 0;
/* Wait status */
int waitStatus;
/* Child is running */
int isRunning = 0;
int exceededLoad = 0;
int exceededCPU = 0;
/* Sleep timespec */
struct timespec sleepTime;
/* Time now and before */
struct timeval now;
struct timeval lastUpdate;
/* CPU usage now and before, plus the limit */
unsigned long cpuNow = 0;
unsigned long cpuLast = 0 ;
unsigned long cpuLimit = 0;
/* Load limit */
float loadLimit = 0.00;
/* bucket holding how many ticks we can consume */
float tickBucket = 0;
/* Setup default sleep period */
unsigned long sleep_us = DEFAULT_SLEEP;
/* Misc */
int i;
/* Our long options */
struct option long_options[] = {
{"pid",0,0,'p'},
{"cpu-limit",0,0,'c'},
{"load-limit",0,0,'l'},
{"verbose",0,0,'v'},
{"help",0,0,'h'},
{0,0,0,0}
};
/* Setup signals */
signal(SIGUSR1, sigusr1);
signal(SIGUSR2, sigusr2);
// signal(SIGCHLD, sigchld);
signal(SIGINT, signal_handler);
signal(SIGHUP, signal_handler);
signal(SIGQUIT, signal_handler);
signal(SIGILL, signal_handler);
signal(SIGKILL, signal_handler);
signal(SIGABRT, signal_handler);
signal(SIGTERM, signal_handler);
signal(SIGPIPE, signal_handler);
signal(SIGSEGV, signal_handler);
/* Loop with options */
while (1) {
int option_index = 0;
char c;
/* Process */
c = getopt_long(argc,argv,"p:c:l:vh",long_options,&option_index);
if (c == -1)
break;
/* Check... */
switch (c) {
case 'p':
pid = atoi(optarg);
break;
case 'c':
cpuLimit = atoi(optarg);
break;
case 'l':
loadLimit = atof(optarg);
break;
case 'v':
verbose++;
break;
case 'h':
printUsage(argv);
return 0;
default:
fprintf(stderr,"Try --help for more info\n");
return 1;
}
}
/* If we don't have a PID we should have a command to run */
if (!pid && optind == argc) {
fprintf(stderr,"%s: Nothing to manage, either --pid/-p or a command must be given\n\n",argv[0]);
fprintf(stderr,"Try --help for more info\n");
return 1;
}
/* If we STILL have params left over, its bad */
if (pid && optind < argc) {
while (optind < argc)
fprintf(stderr,"%s: Invalid option -- %s\n",argv[0],argv[optind++]);
fprintf(stderr,"Try --help for more info\n");
return 1;
}
/* If we don't have a PID, its more than likely we must commandline it */
if (!pid) {
/* Loop with extra args and build our new execve environment */
for (i = 0; i < (argc - optind); i++) {
argv[i] = argv[i + optind];
}
/* End it of with NULL */
argv[i] = NULL;
haveChild = 1;
pid = fork();
if (pid < 0) {
logmsg("ERROR: Failed to fork new process\n");
return 1;
/* Parent */
} else if (pid > 1) {
pgid = pid;
/* Child */
} else {
pid = getpid();
/* Reset parent */
if (setsid () == -1) {
logmsg("ERROR: Error resetting PGID\n");
}
if (verbose) {
// logmsg("Child process %i PRIO set to 20\n",pid);
}
/* FIXME */
// setpriority (PRIO_PROCESS, pid, 20);
execvp(argv[0], argv);
/* We shouldn't really get here */
logmsg("ERROR: Failed to execute command '%s': %s\n",argv[0],strerror(errno));
return 1;
}
/* Set process group if we were specified on the commandline */
} else {
pgid = getpgid(pid);
}
/* Setup child parent group for signalling */
child_pgid = pgid;
/* Continue process */
killpg(pgid, SIGCONT);
isRunning = 1;
/* Last update is right now */
gettimeofday(&lastUpdate,NULL);
if (verbose > 1)
logmsg("Child PID/PGID => %lu/%lu\n",pid,pgid);
/* This is the main program loop */
while (loop) {
/* Period from last check (ms) */
unsigned long elapsed_us;
/* Were we running here? */
int wasRunning = isRunning;
/* Check if we have dead children */
if (haveChild) {
if (waitpid(pid, &waitStatus, WNOHANG) < 0)
break;
/* Check process is still alive */
} else {
if (kill(pid,0) == -1 && errno == ESRCH)
break;
}
/* We need to grab "now" so we can calculate below */
gettimeofday(&now,NULL);
elapsed_us = timediff_us(&now,&lastUpdate);
/* Statistics */
if (!wasRunning)
statsTimeDelayed += elapsed_us;
/* Are we processing load limits? */
if (loadLimit > 0.00) {
double load = getload();
/* Check if our current load is exceeding our limit, stop */
if (load > loadLimit) {
exceededLoad = 1;
/* If we running and we below the threshold, resume */
} else {
exceededLoad = 0;
}
if (verbose > 1)
logmsg("LOAD LIMIT => load %.2f/%.2f\n",load,loadLimit);
}
/* Are we processing cpu limits? */
if (cpuLimit) {
/* Change in ticks for period */
unsigned int ticks_delta;
/* Number of ticks allowed */
double ticks_allowed;
/* Grab current CPU time for entire process group */
cpuNow = getProcessGroupCPUTime(pgid);
/* Change in ticks for period */
ticks_delta = cpuNow - cpuLast;
/* Number of tickes we can eat */
ticks_allowed = ( (float) elapsed_us / (float) CLOCK_PRECISION) * HZ * ( (float) cpuLimit / 100);
/* Remove ticks we ate and add ones we allowed */
tickBucket -= ticks_delta;
tickBucket += ticks_allowed;
/* Check we did not exceed 1s */
if (tickBucket > HZ) {
tickBucket = HZ;
/* Check if we don't have an insane negative value eithr */
} else if (tickBucket < - HZ) {
tickBucket = - HZ;
}
/* If we running and our tick bucket is screwed, stop the process */
if (tickBucket < 0) {
exceededCPU = 1;
sleep_us = DEFAULT_SLEEP;
/* If we not running and we now have some ticks to consume, start the process */
} else if (tickBucket > 0) {
exceededCPU = 0;
/* Set new sleep time */
sleep_us = (tickBucket / HZ) * CLOCK_PRECISION;
};
/* Only use this if its a greater value */
if (sleep_us < DEFAULT_SLEEP)
sleep_us = DEFAULT_SLEEP;
/* If our value is higher than the clock precision we MUST adjust it lower */
/* or we will get a error returned */
else if (sleep_us > CLOCK_PRECISION - DEFAULT_SLEEP)
sleep_us = CLOCK_PRECISION - DEFAULT_SLEEP;
/* Set last values */
cpuLast = cpuNow;
/* Print out info if we're running in verbose mode */
if (verbose > 1)
logmsg("CPU LIMIT => tickBucket = %.2f (allowed += %.2f, consumed -= %lu), elapsed us = %lu, sleep_us = %lu\n", tickBucket, ticks_allowed, ticks_delta, elapsed_us, sleep_us);
}
/* Set last time we were updated */
gettimeofday(&lastUpdate,NULL);
/* If load is high, override the sleep time */
if (exceededLoad) {
sleepTime.tv_sec = 5;
} else {
sleepTime.tv_sec = 0;
}
/* If we running and we should not be, then stop */
if ((exceededLoad || exceededCPU) && wasRunning) {
if (verbose > 2)
logmsg("KILLPG: SIGSTOP sent to %lu\n",pgid);
killpg(pgid,SIGSTOP);
isRunning = 0;
statsSlowdowns++;
/* If we not running and should be then continue */
} else if (!(exceededLoad || exceededCPU) && !wasRunning) {
if (verbose > 2)
logmsg("KILLPG: SIGCONT sent to %lu\n",pgid);
killpg(pgid,SIGCONT);
isRunning = 1;
}
/* Sleep here */
sleepTime.tv_nsec = sleep_us * 1000;
nanosleep(&sleepTime,NULL);
}
if (verbose)
logmsg("STATISTICS: Slowdowns = %lu, Total Time Delayed = %.2fs\n",statsSlowdowns,( (double) statsTimeDelayed / (double) CLOCK_PRECISION ));
return WEXITSTATUS(waitStatus);
}
/* vim: ts=4 */