1 /**
2 * collectd-nagios - src/collectd-nagios.c
3 * Copyright (C) 2008-2010 Florian octo Forster
4 *
5 * Permission is hereby granted, free of charge, to any person obtaining a
6 * copy of this software and associated documentation files (the "Software"),
7 * to deal in the Software without restriction, including without limitation
8 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
9 * and/or sell copies of the Software, and to permit persons to whom the
10 * Software is furnished to do so, subject to the following conditions:
11 *
12 * The above copyright notice and this permission notice shall be included in
13 * all copies or substantial portions of the Software.
14 *
15 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
20 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
21 * DEALINGS IN THE SOFTWARE.
22 *
23 * Authors:
24 * Florian octo Forster <octo at collectd.org>
25 **/
27 #if HAVE_CONFIG_H
28 # include "config.h"
29 #endif
31 #if !defined(__GNUC__) || !__GNUC__
32 # define __attribute__(x) /**/
33 #endif
35 #include <stdlib.h>
36 #include <unistd.h>
37 #include <stdio.h>
38 #include <errno.h>
39 #include <string.h>
40 #include <strings.h>
41 #include <assert.h>
43 #if NAN_STATIC_DEFAULT
44 # include <math.h>
45 /* #endif NAN_STATIC_DEFAULT*/
46 #elif NAN_STATIC_ISOC
47 # ifndef __USE_ISOC99
48 # define DISABLE_ISOC99 1
49 # define __USE_ISOC99 1
50 # endif /* !defined(__USE_ISOC99) */
51 # include <math.h>
52 # if DISABLE_ISOC99
53 # undef DISABLE_ISOC99
54 # undef __USE_ISOC99
55 # endif /* DISABLE_ISOC99 */
56 /* #endif NAN_STATIC_ISOC */
57 #elif NAN_ZERO_ZERO
58 # include <math.h>
59 # ifdef NAN
60 # undef NAN
61 # endif
62 # define NAN (0.0 / 0.0)
63 # ifndef isnan
64 # define isnan(f) ((f) != (f))
65 # endif /* !defined(isnan) */
66 # ifndef isfinite
67 # define isfinite(f) (((f) - (f)) == 0.0)
68 # endif
69 # ifndef isinf
70 # define isinf(f) (!isfinite(f) && !isnan(f))
71 # endif
72 #endif /* NAN_ZERO_ZERO */
74 #include "libcollectdclient/collectd/client.h"
76 #define RET_OKAY 0
77 #define RET_WARNING 1
78 #define RET_CRITICAL 2
79 #define RET_UNKNOWN 3
81 #define CON_NONE 0
82 #define CON_AVERAGE 1
83 #define CON_SUM 2
84 #define CON_PERCENTAGE 3
86 struct range_s
87 {
88 double min;
89 double max;
90 int invert;
91 };
92 typedef struct range_s range_t;
94 extern char *optarg;
95 extern int optind, opterr, optopt;
97 static char *socket_file_g = NULL;
98 static char *value_string_g = NULL;
99 static char *hostname_g = NULL;
101 static range_t range_critical_g;
102 static range_t range_warning_g;
103 static int consolitation_g = CON_NONE;
104 static _Bool nan_is_error_g = 0;
106 static char **match_ds_g = NULL;
107 static size_t match_ds_num_g = 0;
109 /* `strdup' is an XSI extension. I don't want to pull in all of XSI just for
110 * that, so here's an own implementation.. It's easy enough. The GCC attributes
111 * are supposed to get good performance.. -octo */
112 __attribute__((malloc, nonnull (1)))
113 static char *cn_strdup (const char *str) /* {{{ */
114 {
115 size_t strsize;
116 char *ret;
118 strsize = strlen (str) + 1;
119 ret = (char *) malloc (strsize);
120 if (ret != NULL)
121 memcpy (ret, str, strsize);
122 return (ret);
123 } /* }}} char *cn_strdup */
125 static int filter_ds (size_t *values_num,
126 double **values, char ***values_names)
127 {
128 gauge_t *new_values;
129 char **new_names;
131 size_t i;
133 if (match_ds_g == NULL)
134 return (RET_OKAY);
136 new_values = (gauge_t *)calloc (match_ds_num_g, sizeof (*new_values));
137 if (new_values == NULL)
138 {
139 fprintf (stderr, "malloc failed: %s\n", strerror (errno));
140 return (RET_UNKNOWN);
141 }
143 new_names = (char **)calloc (match_ds_num_g, sizeof (*new_names));
144 if (new_names == NULL)
145 {
146 fprintf (stderr, "malloc failed: %s\n", strerror (errno));
147 free (new_values);
148 return (RET_UNKNOWN);
149 }
151 for (i = 0; i < match_ds_num_g; i++)
152 {
153 size_t j;
155 /* match_ds_g keeps pointers into argv but the names will be freed */
156 new_names[i] = cn_strdup (match_ds_g[i]);
157 if (new_names[i] == NULL)
158 {
159 fprintf (stderr, "cn_strdup failed: %s\n", strerror (errno));
160 free (new_values);
161 for (j = 0; j < i; j++)
162 free (new_names[j]);
163 free (new_names);
164 return (RET_UNKNOWN);
165 }
167 for (j = 0; j < *values_num; j++)
168 if (strcasecmp (new_names[i], (*values_names)[j]) == 0)
169 break;
171 if (j == *values_num)
172 {
173 printf ("ERROR: DS `%s' is not available.\n", new_names[i]);
174 free (new_values);
175 for (j = 0; j <= i; j++)
176 free (new_names[j]);
177 free (new_names);
178 return (RET_CRITICAL);
179 }
181 new_values[i] = (*values)[j];
182 }
184 free (*values);
185 for (i = 0; i < *values_num; i++)
186 free ((*values_names)[i]);
187 free (*values_names);
189 *values = new_values;
190 *values_names = new_names;
191 *values_num = match_ds_num_g;
192 return (RET_OKAY);
193 } /* int filter_ds */
195 static void parse_range (char *string, range_t *range)
196 {
197 char *min_ptr;
198 char *max_ptr;
200 if (*string == '@')
201 {
202 range->invert = 1;
203 string++;
204 }
206 max_ptr = strchr (string, ':');
207 if (max_ptr == NULL)
208 {
209 min_ptr = NULL;
210 max_ptr = string;
211 }
212 else
213 {
214 min_ptr = string;
215 *max_ptr = '\0';
216 max_ptr++;
217 }
219 assert (max_ptr != NULL);
221 /* `10' == `0:10' */
222 if (min_ptr == NULL)
223 range->min = 0.0;
224 /* :10 == ~:10 == -inf:10 */
225 else if ((*min_ptr == '\0') || (*min_ptr == '~'))
226 range->min = NAN;
227 else
228 range->min = atof (min_ptr);
230 if ((*max_ptr == '\0') || (*max_ptr == '~'))
231 range->max = NAN;
232 else
233 range->max = atof (max_ptr);
234 } /* void parse_range */
236 static int match_range (range_t *range, double value)
237 {
238 int ret = 0;
240 if (!isnan (range->min) && (range->min > value))
241 ret = 1;
242 if (!isnan (range->max) && (range->max < value))
243 ret = 1;
245 return (((ret - range->invert) == 0) ? 0 : 1);
246 } /* int match_range */
248 static void usage (const char *name)
249 {
250 fprintf (stderr, "Usage: %s <-s socket> <-n value_spec> <-H hostname> [options]\n"
251 "\n"
252 "Valid options are:\n"
253 " -s <socket> Path to collectd's UNIX-socket.\n"
254 " -n <v_spec> Value specification to get from collectd.\n"
255 " Format: `plugin-instance/type-instance'\n"
256 " -d <ds> Select the DS to examine. May be repeated to examine multiple\n"
257 " DSes. By default all DSes are used.\n"
258 " -g <consol> Method to use to consolidate several DSes.\n"
259 " See below for a list of valid arguments.\n"
260 " -H <host> Hostname to query the values for.\n"
261 " -c <range> Critical range\n"
262 " -w <range> Warning range\n"
263 " -m Treat \"Not a Number\" (NaN) as critical (default: warning)\n"
264 "\n"
265 "Consolidation functions:\n"
266 " none: Apply the warning- and critical-ranges to each data-source\n"
267 " individually.\n"
268 " average: Calculate the average of all matching DSes and apply the\n"
269 " warning- and critical-ranges to the calculated average.\n"
270 " sum: Apply the ranges to the sum of all DSes.\n"
271 " percentage: Apply the ranges to the ratio (in percent) of the first value\n"
272 " and the sum of all values."
273 "\n", name);
274 exit (1);
275 } /* void usage */
277 static int do_listval (lcc_connection_t *connection)
278 {
279 lcc_identifier_t *ret_ident = NULL;
280 size_t ret_ident_num = 0;
282 char *hostname = NULL;
284 int status;
285 size_t i;
287 status = lcc_listval (connection, &ret_ident, &ret_ident_num);
288 if (status != 0) {
289 printf ("UNKNOWN: %s\n", lcc_strerror (connection));
290 if (ret_ident != NULL)
291 free (ret_ident);
292 return (RET_UNKNOWN);
293 }
295 status = lcc_sort_identifiers (connection, ret_ident, ret_ident_num);
296 if (status != 0) {
297 printf ("UNKNOWN: %s\n", lcc_strerror (connection));
298 if (ret_ident != NULL)
299 free (ret_ident);
300 return (RET_UNKNOWN);
301 }
303 for (i = 0; i < ret_ident_num; ++i) {
304 char id[1024];
306 if ((hostname_g != NULL) && (strcasecmp (hostname_g, ret_ident[i].host)))
307 continue;
309 if ((hostname == NULL) || strcasecmp (hostname, ret_ident[i].host))
310 {
311 free (hostname);
312 hostname = strdup (ret_ident[i].host);
313 printf ("Host: %s\n", hostname);
314 }
316 /* empty hostname; not to be printed again */
317 ret_ident[i].host[0] = '\0';
319 status = lcc_identifier_to_string (connection,
320 id, sizeof (id), ret_ident + i);
321 if (status != 0) {
322 printf ("ERROR: listval: Failed to convert returned "
323 "identifier to a string: %s\n",
324 lcc_strerror (connection));
325 free (hostname);
326 hostname = NULL;
327 continue;
328 }
330 /* skip over the (empty) hostname and following '/' */
331 printf ("\t%s\n", id + 1);
332 }
334 free (ret_ident);
335 free (hostname);
336 return (RET_OKAY);
337 } /* int do_listval */
339 static int do_check_con_none (size_t values_num,
340 double *values, char **values_names)
341 {
342 int num_critical = 0;
343 int num_warning = 0;
344 int num_okay = 0;
345 const char *status_str = "UNKNOWN";
346 int status_code = RET_UNKNOWN;
347 size_t i;
349 for (i = 0; i < values_num; i++)
350 {
351 if (isnan (values[i]))
352 {
353 if (nan_is_error_g)
354 num_critical++;
355 else
356 num_warning++;
357 }
358 else if (match_range (&range_critical_g, values[i]) != 0)
359 num_critical++;
360 else if (match_range (&range_warning_g, values[i]) != 0)
361 num_warning++;
362 else
363 num_okay++;
364 }
366 if ((num_critical == 0) && (num_warning == 0) && (num_okay == 0))
367 {
368 printf ("WARNING: No defined values found\n");
369 return (RET_WARNING);
370 }
371 else if ((num_critical == 0) && (num_warning == 0))
372 {
373 status_str = "OKAY";
374 status_code = RET_OKAY;
375 }
376 else if (num_critical == 0)
377 {
378 status_str = "WARNING";
379 status_code = RET_WARNING;
380 }
381 else
382 {
383 status_str = "CRITICAL";
384 status_code = RET_CRITICAL;
385 }
387 printf ("%s: %i critical, %i warning, %i okay", status_str,
388 num_critical, num_warning, num_okay);
389 if (values_num > 0)
390 {
391 printf (" |");
392 for (i = 0; i < values_num; i++)
393 printf (" %s=%f;;;;", values_names[i], values[i]);
394 }
395 printf ("\n");
397 return (status_code);
398 } /* int do_check_con_none */
400 static int do_check_con_average (size_t values_num,
401 double *values, char **values_names)
402 {
403 size_t i;
404 double total;
405 int total_num;
406 double average;
407 const char *status_str = "UNKNOWN";
408 int status_code = RET_UNKNOWN;
410 total = 0.0;
411 total_num = 0;
412 for (i = 0; i < values_num; i++)
413 {
414 if (isnan (values[i]))
415 {
416 if (!nan_is_error_g)
417 continue;
419 printf ("CRITICAL: Data source \"%s\" is NaN\n",
420 values_names[i]);
421 return (RET_CRITICAL);
422 }
424 total += values[i];
425 total_num++;
426 }
428 if (total_num == 0)
429 {
430 printf ("WARNING: No defined values found\n");
431 return (RET_WARNING);
432 }
434 average = total / total_num;
436 if (match_range (&range_critical_g, average) != 0)
437 {
438 status_str = "CRITICAL";
439 status_code = RET_CRITICAL;
440 }
441 else if (match_range (&range_warning_g, average) != 0)
442 {
443 status_str = "WARNING";
444 status_code = RET_WARNING;
445 }
446 else
447 {
448 status_str = "OKAY";
449 status_code = RET_OKAY;
450 }
452 printf ("%s: %g average |", status_str, average);
453 for (i = 0; i < values_num; i++)
454 printf (" %s=%f;;;;", values_names[i], values[i]);
455 printf ("\n");
457 return (status_code);
458 } /* int do_check_con_average */
460 static int do_check_con_sum (size_t values_num,
461 double *values, char **values_names)
462 {
463 size_t i;
464 double total;
465 int total_num;
466 const char *status_str = "UNKNOWN";
467 int status_code = RET_UNKNOWN;
469 total = 0.0;
470 total_num = 0;
471 for (i = 0; i < values_num; i++)
472 {
473 if (isnan (values[i]))
474 {
475 if (!nan_is_error_g)
476 continue;
478 printf ("CRITICAL: Data source \"%s\" is NaN\n",
479 values_names[i]);
480 return (RET_CRITICAL);
481 }
483 total += values[i];
484 total_num++;
485 }
487 if (total_num == 0)
488 {
489 printf ("WARNING: No defined values found\n");
490 return (RET_WARNING);
491 }
493 if (match_range (&range_critical_g, total) != 0)
494 {
495 status_str = "CRITICAL";
496 status_code = RET_CRITICAL;
497 }
498 else if (match_range (&range_warning_g, total) != 0)
499 {
500 status_str = "WARNING";
501 status_code = RET_WARNING;
502 }
503 else
504 {
505 status_str = "OKAY";
506 status_code = RET_OKAY;
507 }
509 printf ("%s: %g sum |", status_str, total);
510 for (i = 0; i < values_num; i++)
511 printf (" %s=%f;;;;", values_names[i], values[i]);
512 printf ("\n");
514 return (status_code);
515 } /* int do_check_con_sum */
517 static int do_check_con_percentage (size_t values_num,
518 double *values, char **values_names)
519 {
520 size_t i;
521 double sum = 0.0;
522 double percentage;
524 const char *status_str = "UNKNOWN";
525 int status_code = RET_UNKNOWN;
527 if ((values_num < 1) || (isnan (values[0])))
528 {
529 printf ("WARNING: The first value is not defined\n");
530 return (RET_WARNING);
531 }
533 for (i = 0; i < values_num; i++)
534 {
535 if (isnan (values[i]))
536 {
537 if (!nan_is_error_g)
538 continue;
540 printf ("CRITICAL: Data source \"%s\" is NaN\n",
541 values_names[i]);
542 return (RET_CRITICAL);
543 }
545 sum += values[i];
546 }
548 if (sum == 0.0)
549 {
550 printf ("WARNING: Values sum up to zero\n");
551 return (RET_WARNING);
552 }
554 percentage = 100.0 * values[0] / sum;
556 if (match_range (&range_critical_g, percentage) != 0)
557 {
558 status_str = "CRITICAL";
559 status_code = RET_CRITICAL;
560 }
561 else if (match_range (&range_warning_g, percentage) != 0)
562 {
563 status_str = "WARNING";
564 status_code = RET_WARNING;
565 }
566 else
567 {
568 status_str = "OKAY";
569 status_code = RET_OKAY;
570 }
572 printf ("%s: %lf percent |", status_str, percentage);
573 for (i = 0; i < values_num; i++)
574 printf (" %s=%lf;;;;", values_names[i], values[i]);
575 return (status_code);
576 } /* int do_check_con_percentage */
578 static int do_check (lcc_connection_t *connection)
579 {
580 gauge_t *values;
581 char **values_names;
582 size_t values_num;
583 char ident_str[1024];
584 lcc_identifier_t ident;
585 size_t i;
586 int status;
588 snprintf (ident_str, sizeof (ident_str), "%s/%s",
589 hostname_g, value_string_g);
590 ident_str[sizeof (ident_str) - 1] = 0;
592 memset (&ident, 0, sizeof (ident));
593 status = lcc_string_to_identifier (connection, &ident, ident_str);
594 if (status != 0)
595 {
596 printf ("ERROR: Creating an identifier failed: %s.\n",
597 lcc_strerror (connection));
598 LCC_DESTROY (connection);
599 return (RET_CRITICAL);
600 }
602 status = lcc_getval (connection, &ident,
603 &values_num, &values, &values_names);
604 if (status != 0)
605 {
606 printf ("ERROR: Retrieving values from the daemon failed: %s.\n",
607 lcc_strerror (connection));
608 LCC_DESTROY (connection);
609 return (RET_CRITICAL);
610 }
612 LCC_DESTROY (connection);
614 status = filter_ds (&values_num, &values, &values_names);
615 if (status != RET_OKAY)
616 return (status);
618 status = RET_UNKNOWN;
619 if (consolitation_g == CON_NONE)
620 status = do_check_con_none (values_num, values, values_names);
621 else if (consolitation_g == CON_AVERAGE)
622 status = do_check_con_average (values_num, values, values_names);
623 else if (consolitation_g == CON_SUM)
624 status = do_check_con_sum (values_num, values, values_names);
625 else if (consolitation_g == CON_PERCENTAGE)
626 status = do_check_con_percentage (values_num, values, values_names);
628 free (values);
629 if (values_names != NULL)
630 for (i = 0; i < values_num; i++)
631 free (values_names[i]);
632 free (values_names);
634 return (status);
635 } /* int do_check */
637 int main (int argc, char **argv)
638 {
639 char address[1024];
640 lcc_connection_t *connection;
642 int status;
644 range_critical_g.min = NAN;
645 range_critical_g.max = NAN;
646 range_critical_g.invert = 0;
648 range_warning_g.min = NAN;
649 range_warning_g.max = NAN;
650 range_warning_g.invert = 0;
652 while (42)
653 {
654 int c;
656 c = getopt (argc, argv, "w:c:s:n:H:g:d:hm");
657 if (c < 0)
658 break;
660 switch (c)
661 {
662 case 'c':
663 parse_range (optarg, &range_critical_g);
664 break;
665 case 'w':
666 parse_range (optarg, &range_warning_g);
667 break;
668 case 's':
669 socket_file_g = optarg;
670 break;
671 case 'n':
672 value_string_g = optarg;
673 break;
674 case 'H':
675 hostname_g = optarg;
676 break;
677 case 'g':
678 if (strcasecmp (optarg, "none") == 0)
679 consolitation_g = CON_NONE;
680 else if (strcasecmp (optarg, "average") == 0)
681 consolitation_g = CON_AVERAGE;
682 else if (strcasecmp (optarg, "sum") == 0)
683 consolitation_g = CON_SUM;
684 else if (strcasecmp (optarg, "percentage") == 0)
685 consolitation_g = CON_PERCENTAGE;
686 else
687 {
688 fprintf (stderr, "Unknown consolidation function `%s'.\n",
689 optarg);
690 usage (argv[0]);
691 }
692 break;
693 case 'd':
694 {
695 char **tmp;
696 tmp = (char **) realloc (match_ds_g,
697 (match_ds_num_g + 1)
698 * sizeof (char *));
699 if (tmp == NULL)
700 {
701 fprintf (stderr, "realloc failed: %s\n",
702 strerror (errno));
703 return (RET_UNKNOWN);
704 }
705 match_ds_g = tmp;
706 match_ds_g[match_ds_num_g] = cn_strdup (optarg);
707 if (match_ds_g[match_ds_num_g] == NULL)
708 {
709 fprintf (stderr, "cn_strdup failed: %s\n",
710 strerror (errno));
711 return (RET_UNKNOWN);
712 }
713 match_ds_num_g++;
714 break;
715 }
716 case 'm':
717 nan_is_error_g = 1;
718 break;
719 default:
720 usage (argv[0]);
721 } /* switch (c) */
722 }
724 if ((socket_file_g == NULL) || (value_string_g == NULL)
725 || ((hostname_g == NULL) && (strcasecmp (value_string_g, "LIST"))))
726 {
727 fprintf (stderr, "Missing required arguments.\n");
728 usage (argv[0]);
729 }
731 snprintf (address, sizeof (address), "unix:%s", socket_file_g);
732 address[sizeof (address) - 1] = 0;
734 connection = NULL;
735 status = lcc_connect (address, &connection);
736 if (status != 0)
737 {
738 printf ("ERROR: Connecting to daemon at %s failed.\n",
739 socket_file_g);
740 return (RET_CRITICAL);
741 }
743 if (0 == strcasecmp (value_string_g, "LIST"))
744 return (do_listval (connection));
746 return (do_check (connection));
747 } /* int main */