1 <!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook V4.1//EN">
2 <book>
3 <title>Nagios Plug-in Developer Guidelines</title>
5 <bookinfo>
6 <authorgroup>
7 <author>
8 <firstname>Karl</firstname>
9 <surname>DeBisschop</surname>
10 <affiliation>
11 <address><email>karl@debisschop.net</email></address>
12 </affiliation>
13 </author>
15 <author>
16 <firstname>Ethan</firstname>
17 <surname>Galstad</surname>
18 <authorblurb>
19 <para>Author of Nagios</para>
20 <para><ulink url="http://www.nagios.org"></ulink></para>
21 </authorblurb>
22 <affiliation>
23 <address><email>netsaint@linuxbox.com</email></address>
24 </affiliation>
25 </author>
27 <author>
28 <firstname>Hugo</firstname>
29 <surname>Gayosso</surname>
30 <affiliation>
31 <address><email>hgayosso@gnu.org</email></address>
32 </affiliation>
33 </author>
36 <author>
37 <firstname>Subhendu</firstname>
38 <surname>Ghosh</surname>
39 <affiliation>
40 <address><email>sghosh@sourceforge.net</email></address>
41 </affiliation>
42 </author>
44 <author>
45 <firstname>Stanley</firstname>
46 <surname>Hopcroft</surname>
47 <affiliation>
48 <address><email>stanleyhopcroft@sourceforge.net</email></address>
49 </affiliation>
50 </author>
52 </authorgroup>
54 <pubdate>2002</pubdate>
55 <title>Nagios plug-in development guidelines</title>
57 <revhistory>
58 <revision>
59 <revnumber>0.4</revnumber>
60 <date>2 May 2002</date>
61 </revision>
62 </revhistory>
64 <copyright>
65 <year>2000 2001 2002</year>
66 <holder>Karl DeBisschop, Ethan Galstad,
67 Hugo Gayosso, Stanley Hopcroft, Subhendu Ghosh</holder>
68 </copyright>
70 </bookinfo>
73 <preface id=preface>
74 <title>About the guidelines</title>
76 <para>The purpose of this guidelines is to provide a reference for
77 the plug-in developers and encourage the standarization of the
78 different kind of plug-ins: C, shell, perl, python, etc.</para>
81 <section> <title>Copyright</title>
83 <para>Nagios Plug-in Development Guidelines Copyright (C) 2000 2001
84 2002
85 Karl DeBisschop, Ethan Galstad, Hugo Gayosso, Stanley Hopcroft,
86 Subhendu Ghosh</para>
88 <para>Permission is granted to make and distribute verbatim
89 copies of this manual provided the copyright notice and this
90 permission notice are preserved on all copies.</para>
92 <para>The plugins themselves are copyrighted by their respective
93 authors.</para>
95 </section>
96 </preface>
98 <article>
99 <section id="PlugOutput"><title>Plugin Output for Nagios</title>
101 <para>You should always print something to STDOUT that tells if the
102 service is working or why its failing. Try to keep the output short -
103 probably less that 80 characters. Remember that you ideally would like
104 the entire output to appear in a pager message, which will get chopped
105 off after a certain length.</para>
107 <section><title>Print only one line of text</title>
108 <para>Nagios will only grab the first line of text from STDOUT
109 when it notifies contacts about potential problems. If you print
110 multiple lines, you're out of luck. Remember, keep it short and
111 to the point.</para>
112 </section>
114 <section><title>Verbose output</title>
115 <para>Use the -v flag for verbose output. You should allow multiple
116 -v options for additional verbosity, up to a maximum of 3. The standard
117 type of output should be:</para>
119 <table id="verbose_levels"><title>Verbose output levels</title>
120 <tgroup cols="2">
121 <thead>
122 <row>
123 <entry><para>Verbosity level</para></entry>
124 <entry><para>Type of output</para></entry>
125 </row>
126 </thead>
127 <tbody>
128 <row>
129 <entry align=center><para>0</para></entry>
130 <entry><para>Single line, minimal output. Summary</para></entry>
131 </row>
132 <row>
133 <entry align=center><para>1</para></entry>
134 <entry><para>Single line, additional information (eg list processes that fail)</para></entry>
135 </row>
136 <row>
137 <entry align=center><para>2</para></entry>
138 <entry><para>Multi line, configuration debug output (eg ps command used)</para></entry>
139 </row>
140 <row>
141 <entry align=center><para>3</para></entry>
142 <entry><para>Lots of detail for plugin problem diagnosis</para></entry>
143 </row>
144 </tbody>
145 </tgroup>
146 </table>
147 </section>
149 <section><title>Screen Output</title>
150 <para>The plug-in should print the diagnostic and just the
151 synopsis part of the help message. A well written plugin would
152 then have --help as a way to get the verbose help.</para>
153 <para>Code and output should try to respect the 80x25 size of a
154 crt (remember when fixing stuff in the server room!)</para>
155 </section>
157 <section><title>Return the proper status code</title>
158 <para>See <xref linkend="ReturnCodes"> below
159 for the numeric values of status codes and their
160 description. Remember to return an UNKNOWN state if bogus or
161 invalid command line arguments are supplied or it you are unable
162 to check the service.</para>
163 </section>
165 <section><title>Plugin Return Codes</title>
166 <para>The return codes below are based on the POSIX spec of returning
167 a positive value. Netsaint prior to v0.0.7 supported non-POSIX
168 compliant return code of "-1" for unknown. Nagios supports POSIX return
169 codes by default.</para>
171 <para>Note: Some plugins will on occasion print on STDOUT that an error
172 occurred and error code is 138 or 255 or some such number. These
173 are usually caused by plugins using system commands and having not
174 enough checks to catch unexpected output. Developers should include a
175 default catch-all for system command output that returns an UNKOWN
176 return code.</para>
178 <table id="ReturnCodes"><title>Plugin Return Codes</title>
179 <tgroup cols="3">
180 <thead>
181 <row>
182 <entry><para>Numeric Value</para></entry>
183 <entry><para>Service Status</para></entry>
184 <entry><para>Status Description</para></entry>
185 </row>
186 </thead>
187 <tbody>
188 <row>
189 <entry align=center><para>0</para></entry>
190 <entry valign=middle><para>OK</para></entry>
191 <entry><para>The plugin was able to check the service and it
192 appeared to be functioning properly</para></entry>
193 </row>
194 <row>
195 <entry align=center><para>1</para></entry>
196 <entry valign=middle><para>Warning</para></entry>
197 <entry><para>The plugin was able to check the service, but it
198 appeared to be above some "warning" threshold or did not appear
199 to be working properly</para></entry>
200 </row>
201 <row>
202 <entry align=center><para>2</para></entry>
203 <entry valign=middle><para>Critical</para></entry>
204 <entry><para>The plugin detected that either the service was not
205 running or it was above some "critical" threshold</para></entry>
206 </row>
207 <row>
208 <entry align=center><para>3</para></entry>
209 <entry valign=middle><para>Unknown</para></entry>
210 <entry><para>Invalid command line arguments were supplied to the
211 plugin or the plugin was unable to check the status of the given
212 hosts/service</para></entry>
213 </row>
214 </tbody>
215 </tgroup>
216 </table>
219 </section>
222 </section>
224 <section id="SysCmdAuxFiles"><title>System Commands and Auxiliary Files</title>
226 <section><title>Don't execute system commands without specifying their
227 full path</title>
228 <para>Don't use exec(), popen(), etc. to execute external
229 commands without explicity using the full path of the external
230 program.</para>
232 <para>Doing otherwise makes the plugin vulnerable to hijacking
233 by a trojan horse earlier in the search path. See the main
234 plugin distribution for examples on how this is done.</para>
235 </section>
237 <section><title>Use spopen() if external commands must be executed</title>
239 <para>If you have to execute external commands from within your
240 plugin and you're writing it in C, use the spopen() function
241 that Karl DeBisschop has written.</para>
243 <para>The code for spopen() and spclose() is included with the
244 core plugin distribution.</para>
245 </section>
247 <section><title>Don't make temp files unless absolutely required</title>
249 <para>If temp files are needed, make sure that the plugin will
250 fail cleanly if the file can't be written (e.g., too few file
251 handles, out of disk space, incorrect permissions, etc.) and
252 delete the temp file when processing is complete.</para>
253 </section>
255 <section><title>Don't be tricked into following symlinks</title>
257 <para>If your plugin opens any files, take steps to ensure that
258 you are not following a symlink to another location on the
259 system.</para>
260 </section>
262 <section><title>Validate all input</title>
264 <para>use routines in utils.c or utils.pm and write more as needed</para>
265 </section>
267 </section>
272 <section id="PerlPlugin"><title>Perl Plugins</title>
274 <para>Perl plugins are coded a little more defensively than other
275 plugins because of embedded Perl. When configured as such, embedded
276 Perl Nagios (ePN) requires stricter use of the some of Perl's features.
277 This section outlines some of the steps needed to use ePN
278 effectively.</para>
280 <orderedlist>
282 <listitem><para> Do not use BEGIN and END blocks since they will be called
283 the first time and when Nagios shuts down with Embedded Perl (ePN). In
284 particular, do not use BEGIN blocks to initialize variables.</para>
285 </listitem>
287 <listitem><para>To use utils.pm, you need to provide a full path to the
288 module in order for it to work with ePN.</para>
290 <literallayout>
291 e.g.
292 use lib "/usr/local/nagios/libexec";
293 use utils qw(...);
294 </literallayout>
295 </listitem>
297 <listitem><para>Perl scripts should be called with "-w"</para>
298 </listitem>
300 <listitem><para>All Perl plugins must compile cleanly under "use strict" - i.e. at
301 least explicitly package names as in "$main::x" or predeclare every
302 variable. </para>
305 <para>Explicitly initialize each varialable in use. Otherwise with
306 caching enabled, the plugin will not be recompilied each time, and
307 therefore Perl will not reinitialize all the variables. All old
308 variable values will still be in effect.</para>
309 </listitem>
311 <listitem><para>Do not use < DATA > (these simply do not compile under ePN).</para>
312 </listitem>
314 <listitem><para>Do not use named subroutines</para>
315 </listitem>
317 <listitem><para>If writing to a file (perhaps recording
318 performance data) explicitly close close it. The plugin never
319 calls <emphasis role=strong>exit</emphasis>; that is caught by
320 p1.pl, so output streams are never closed.</para>
321 </listitem>
323 <listitem><para>As in <xref linkend="runtime"> all plugins need
324 to monitor their runtime, specially if they are using network
325 resources. Use of the <emphasis>alarm</emphasis> is recommended.
326 Plugins may import a default time out ($TIMEOUT) from utils.pm.
327 </para>
328 </listitem>
330 <listitem><para>Perl plugins should import %ERRORS from utils.pm
331 and then "exit $ERRORS{'OK'}" rather than "exit 0"
332 </para>
333 </listitem>
335 </orderedlist>
337 </section>
339 <section id="runtime"><title>Runtime Timeouts</title>
341 <para>Plugins have a very limited runtime - typically 10 sec.
342 As a result, it is very important for plugins to maintain internal
343 code to exit if runtime exceeds a threshold. </para>
345 <para>All plugins should timeout gracefully, not just networking
346 plugins. For instance, df may lock if you have automounted
347 drives and your network fails - but on first glance, who'd think
348 df could lock up like that. Plus, it should just be more error
349 resistant to be able to time out rather than consume
350 resources.</para>
352 <section><title>Use DEFAULT_SOCKET_TIMEOUT</title>
354 <para>All network plugins should use DEFAULT_SOCKET_TIMEOUT to timeout</para>
356 </section>
359 <section><title>Add alarms to network plugins</title>
361 <para>If you write a plugin which communicates with another
362 networked host, you should make sure to set an alarm() in your
363 code that prevents the plugin from hanging due to abnormal
364 socket closures, etc. Nagios takes steps to protect itself
365 against unruly plugins that timeout, but any plugins you create
366 should be well behaved on their own.</para>
368 </section>
372 </section>
374 <section id="PlugOptions"><title>Plugin Options</title>
376 <para>A well written plugin should have --help as a way to get
377 verbose help. Code and output should try to respect the 80x25 size of a
378 crt (remember when fixing stuff in the server room!)</para>
380 <section><title>Option Processing</title>
382 <para>For plugins written in C, we recommend the C standard
383 getopt library for short options. Getopt_long is always available.
384 </para>
386 <para>For plugins written in Perl, we recommend Getopt::Long module.</para>
388 <para>Positional arguments are strongly discouraged.</para>
390 <para>There are a few reserved options that should not be used
391 for other purposes:</para>
393 <literallayout>
394 -V version (--version)
395 -h help (--help)
396 -t timeout (--timeout)
397 -w warning threshold (--warning)
398 -c critical threshold (--critical)
399 -H hostname (--hostname)
400 -v verbose (--verbose)
401 </literallayout>
403 <para>In addition to the reserved options above, some other standard options are:</para>
405 <literallayout>
406 -C SNMP community (--community)
407 -a authentication password (--authentication)
408 -l login name (--logname)
409 -p port or password (--port or --passwd/--password)monitors operational
410 -u url or username (--url or --username)
411 </literallayout>
413 <para>Look at check_pgsql and check_procs to see how I currently
414 think this can work. Standard options are:</para>
417 <para>The option -V or --version should be present in all
418 plugins. For C plugins it should result in a call to print_revision, a
419 function in utils.c which takes two character arguments, the
420 command name and the plugin revision.</para>
422 <para>The -? option, or any other unparsable set of options,
423 should print out a short usage statement. Character width should
424 be 80 and less and no more that 23 lines should be printed (it
425 should display cleanly on a dumb terminal in a server
426 room).</para>
428 <para>The option -h or --help should be present in all plugins.
429 In C plugins, it should result in a call to print_help (or
430 equivalent). The function print_help should call print_revision,
431 then print_usage, then should provide detailed
432 help. Help text should fit on an 80-character width display, but
433 may run as many lines as needed.</para>
435 <para>The option -v or --verbose should be present in all plugins.
436 The user should be allowed to specify -v multiple times to increase
437 the verbosity level, as described in <xref linkend="verbose_levels">.</para>
438 </section>
440 <section>
441 <title>Plugins with more than one type of threshold, or with
442 threshold ranges</title>
444 <para>Old style was to do things like -ct for critical time and
445 -cv for critical value. That goes out the window with POSIX
446 getopt. The allowable alternatives are:</para>
448 <orderedlist>
449 <listitem>
450 <para>long options like -critical-time (or -ct and -cv, I
451 suppose).</para>
452 </listitem>
454 <listitem>
455 <para>repeated options like `check_load -w 10 -w 6 -w 4 -c
456 16 -c 10 -c 10`</para>
457 </listitem>
459 <listitem>
460 <para>for brevity, the above can be expressed as `check_load
461 -w 10,6,4 -c 16,10,10`</para>
462 </listitem>
464 <listitem>
465 <para>ranges are expressed with colons as in `check_procs -C
466 httpd -w 1:20 -c 1:30` which will warn above 20 instances,
467 and critical at 0 and above 30</para>
468 </listitem>
470 <listitem>
471 <para>lists are expressed with commas, so Jacob's check_nmap
472 uses constructs like '-p 1000,1010,1050:1060,2000'</para>
473 </listitem>
475 <listitem>
476 <para>If possible when writing lists, use tokens to make the
477 list easy to remember and non-order dependent - so
478 check_disk uses '-c 10000,10%' so that it is clear which is
479 the precentage and which is the KB values (note that due to
480 my own lack of foresight, that used to be '-c 10000:10%' but
481 such constructs should all be changed for consistency,
482 though providing reverse compatibility is fairly
483 easy).</para>
484 </listitem>
486 </orderedlist>
488 <para>As always, comments are welcome - making this consistent
489 without a host of long options was quite a hassle, and I would
490 suspect that there are flaws in this strategy.
491 </para>
492 </section>
493 </section>
495 <section id="CodingGuidelines"><title>Coding guidelines</title>
496 <para>See <ulink url="http://www.gnu.org/prep/standards_toc.html">GNU
497 Coding standards</ulink> for general guidelines.</para>
498 <section><title>Comments</title>
499 <para>You should use /* */ for comments and not // as some compilers
500 do not handle the latter form.</para>
501 </section>
503 <section><title>CVS comments</title>
504 <para>When adding CVS comments at commit time, you can use the following prefixes:
505 <variablelist>
506 <varlistentry><term>- comment</term>
507 <listitem>
508 <para>for a comment that can be removed from the Changelog</para>
509 </listitem>
510 </varlistentry>
511 <varlistentry><term>* comment</term>
512 <listitem>
513 <para>for an important amendment to be included into a features list</para>
514 </listitem>
515 </varlistentry>
516 </variablelist>
517 </para>
518 </section>
519 </section>
521 <section id="SubmittingChanges"><title>Submission of new plugins and patches</title>
523 <section id="Patches"><title>Patches</title>
524 <para>If you have a bug patch, please supply a unified or context diff against the
525 version you are using. For new features, please supply a diff against
526 the CVS HEAD version.</para>
528 <para>Patches should be submitted via
529 <ulink url="http://sourceforge.net/tracker/?group_id=29880&atid=397599">SourceForge's
530 tracker system for Nagiosplug patches</ulink>
531 and be announced to the nagiosplug-devel mailing list.</para>
532 </section>
534 <section id="New_plugins"><title>New plugins</title>
535 <para>If you would like others to use your plugins and have it included in
536 the standard distribution, please include patches for the relevant
537 configuration files, in particular "configure.in". Otherwise submitted
538 plugins will be included in the contrib directory.</para>
540 <para>Plugins in the contrib directory are going to be migrated to the
541 standard plugins/plugin-scripts directory as time permits and per user
542 requests. The minimum requirements are:</para>
544 <orderedlist>
545 <listitem>
546 <para>The standard command options are supported (--help, --version,
547 --timeout, --warning, --critical)</para>
548 </listitem>
549 <listitem>
550 <para>It is determined to be not redundant (for instance, we would not
551 add a new version of check_disk just because someone had provide
552 a plugin that had perf checking - we would incorporate the features
553 into an exisiting plugin)</para>
554 </listitem>
555 <listitem>
556 <para>One of the developers has had the time to audit the code and declare
557 it ready for core</para>
558 </listitem>
559 <listitem>
560 <para>It should also follow code format guidelines, and use functions from
561 utils (perl or c or sh) rather than cooking it's own</para>
562 </listitem>
563 </orderedlist>
565 <para>New plugins should be submitted via
566 <ulink url="http://sourceforge.net/tracker/?group_id=29880&atid=541465">SourceForge's
567 tracker system for Nagiosplug new plugins</ulink>
568 and be announced to the nagiosplug-devel mailing list.</para>
570 <para>For new plugins, provide a diff to add to the EXTRAS list (configure.in)
571 unless you are fairly sure that the plugin will work for all platforms with
572 no non-standard software added.</para>
574 <para>If possible please submit a test harness. Documentation on sample
575 tests coming soon.</para>
576 </section>
578 </section>
580 </article>
582 </book>