1 collectd - System information collection daemon
2 =================================================
8 collectd is a small daemon which collects system information periodically
9 and provides mechanisms to store and monitor the values in a variety of
16 * collectd is able to collect the following data:
19 Apache server utilization: Number of bytes transfered, number of
20 requests handled and detailed scoreboard statistics
23 APC UPS Daemon: UPS charge, load, input/output/battery voltage, etc.
26 Sensors in Macs running Mac OS X / Darwin: Temperature, fanspeed and
30 Statistics about Ascent, a free server for the game `World of Warcraft'.
33 Batterycharge, -current and voltage of ACPI and PMU based laptop
37 CPU utilization: Time spent in the system, user, nice, idle, and related
41 CPU frequency (For laptops with speed step or a similar technology)
44 Mountpoint usage (Basically the values `df(1)' delivers)
47 Disk utilization: Sectors read/written, number of read/write actions,
48 average time an IO-operation took to complete.
51 DNS traffic: Query types, response codes, opcodes and traffic/octets
55 Email statistics: Count, traffic, spam scores and checks.
56 See collectd-email(5).
59 Amount of entropy available to the system.
62 Values gathered by a custom program or script.
66 Harddisk temperatures using hddtempd.
69 Interface traffic: Number of octets, packets and errors for each
73 Iptables' counters: Number of bytes that were matched by a certain
77 IPVS connection statistics (number of connections, octets and packets
78 for each service and destination).
79 See http://www.linuxvirtualserver.org/software/index.html.
82 IRQ counters: Frequency in which certain interrupts occur.
85 System load average over the last 1, 5 and 15 minutes.
88 CPU, disk and network I/O statistics from virtual machines.
91 Motherboard sensors: temperature, fanspeed and voltage information,
95 Statistics of the memcached distributed caching system.
96 <http://www.danga.com/memcached/>
99 Memory utilization: Memory occupied by running processes, page cache,
100 buffer cache and free.
103 Information provided by serial multimeters, such as the `Metex
107 MySQL server statistics: Commands issued, handlers triggered, thread
108 usage, query cache utilization and traffic/octets sent and received.
111 Very detailed Linux network interface and routing statistics. You can get
112 (detailed) information on interfaces, qdiscs, classes, and, if you can
113 make use of it, filters.
116 Receive values that were collected by other hosts. Large setups will
117 want to collect the data on one dedicated machine, and this is the
118 plugin of choice for that.
121 NFS Procedures: Which NFS command were called how often. Only NFSv2 and
125 Collects statistics from `nginx' (speak: engine X), a HTTP and mail
129 NTP daemon statistics: Local clock drift, offset to peers, etc.
132 Network UPS tools: UPS current, voltage, power, charge, utilisation,
133 temperature, etc. See upsd(8).
136 The perl plugin implements a Perl-interpreter into collectd. You can
137 write your own plugins in Perl and return arbitrary values using this
138 API. See collectd-perl(5).
141 Network latency: Time to reach the default gateway or another given
145 Process counts: Number of running, sleeping, zombie, ... processes.
148 System sensors, accessed using lm_sensors: Voltages, temperatures and
152 RX and TX of serial interfaces. Linux only; needs root privileges.
155 Read values from SNMP (Simple Network Management Protocol) enabled
156 network devices such as switches, routers, thermometers, rack monitoring
157 servers, etc. See collectd-snmp(5).
160 Pages swapped out onto harddisk or whatever is called `swap' by the OS..
163 Follows (tails) logfiles, parses them by lines and submits matched
167 Bytes and operations read and written on tape devices. Solaris only.
170 Number of TCP connections to specific local and remote ports.
173 Users currently logged in.
176 Virtual memory statistics, e. g. the number of page-ins/-outs or the
177 number of pagefaults.
180 System resources used by Linux VServers.
181 See <http://linux-vserver.org/>.
184 Link quality of wireless cards. Linux only.
187 Bitrate and frequency of music played with XMMS.
189 * Output can be written or send to various destinations by the following
193 Write to comma separated values (CSV) files. This needs lots of
194 diskspace but is extremely portable and can be analysed with almost
195 every program that can analyse anything. Even Microsoft's Excel..
198 Send the data to a remote host to save the data somehow. This is useful
199 for large setups where the data should be saved by a dedicated machine.
202 Of course the values are propagated to plugins written in Perl, too, so
203 you can easily do weird stuff with the plugins we didn't dare think of
204 ;) See collectd-perl(5).
207 Output to round-robin-database (RRD) files using librrd. See rrdtool(1).
208 This is likely the most popular destination for such values. Since
209 updates to RRD-files are somewhat expensive this plugin can cache
210 updates to the files and write a bunch of updates at once, which lessens
214 One can query the values from the unixsock plugin whenever they're
215 needed. Please read collectd-unixsock(5) for a description on how that's
218 * Logging is, as everything in collectd, provided by plugins. The following
219 plugins keep up informed about what's going on:
222 Writes logmessages to a file or STDOUT/STDERR.
225 Log messages are propagated to plugins written in Perl as well.
226 See collectd-perl(5).
229 Logs to the standard UNIX logging mechanism, syslog.
231 * Notifications can be handled by the following plugins:
234 Execute a program or script to handle the notification.
235 See collectd-exec(5).
238 Writes the notification message to a file or STDOUT/STDERR.
241 Send the notification to a remote host to handle it somehow.
244 Notifications are propagated to plugins written in Perl as well.
245 See collectd-perl(5).
247 * Miscellaneous plugins:
250 Sets the hostname to an unique identifier. This is meant for setups
251 where each client may migrate to another physical host, possibly going
252 through one or more name changes in the process.
254 * Performance: Since collectd is running as a daemon it doesn't spend much
255 time starting up again and again. With the exception of the exec plugin no
256 processes are forked. Caching in output plugins, such as the rrdtool and
257 network plugins, makes sure your resources are used efficiently. Also,
258 since collectd is programmed multithreaded it benefits from hyperthreading
259 and multicore processors and makes sure that the daemon isn't idle if only
260 one plugins waits for an IO-operation to complete.
262 * Once set up, hardly any maintenance is necessary. Setup is kept as easy
263 as possible and the default values should be okay for most users.
269 * collectd's configuration file can be found at `sysconfdir'/collectd.conf.
270 Run `collectd -h' for a list of builtin defaults. See `collectd.conf(5)'
271 for a list of options and a syntax description.
273 * When the `csv' or `rrdtool' plugins are loaded they'll write the values to
274 files. The usual place for these files is beneath `/var/lib/collectd'.
276 * When using some of the plugins, collectd needs to run as user root, since
277 only root can do certain things, such as craft ICMP packages needed to ping
278 other hosts. collectd should NOT be installed setuid root since it can be
279 used to overwrite valuable files!
281 * Sample scripts to generate graphs reside in `contrib/' in the source
282 package or somewhere near `/usr/share/doc/collectd' in most distributions.
283 Please be aware that those script are meant as a starting point for your
284 own experiments.. Some of them require the `RRDs' Perl module.
285 (`librrds-perl' on Debian) If you have written a more sophisticated
286 solution please share it with us.
288 * The RRAs of the automatically created RRD files depend on the `step'
289 and `heartbeat' settings given. If change these settings you may need to
290 re-create the files, losing all data. Please be aware of that when changing
291 the values and read the rrdtool(1) manpage thoroughly.
294 collectd and chkrootkit
295 -----------------------
297 If you are using the `dns' plugin chkrootkit(1) will report collectd as a
298 packet sniffer ("<iface>: PACKET SNIFFER(/usr/sbin/collectd[<pid>])"). The
299 plugin captures all UDP packets on port 53 to analyze the DNS traffic. In
300 this case, collectd is a legitimate sniffer and the report should be
301 considered to be a false positive. However, you might want to check that
302 this really is collectd and not some other, illegitimate sniffer.
308 To compile collectd from source you will need:
310 * Usual suspects: C compiler, linker, preprocessor, make, ...
312 * A POSIX-threads (pthread) implementation.
313 Since gathering some statistics is slow (network connections, slow devices,
314 etc) the collectd is parallelized. The POSIX threads interface is being
315 used and should be found in various implementations for hopefully all
319 If you want to use the `apache', `ascent', or `nginx' plugin.
322 If present, the uuid plugin will check for UUID from HAL.
325 For querying iptables counters.
327 * libmysqlclient (optional)
328 Unsurprisingly used by the `mysql' plugin.
330 * libnetlink (optional)
331 Used, obviously, for the `netlink' plugin.
333 * libnetsnmp (optional)
334 For the `snmp' plugin.
336 * liboping (optional, if not found a version shipped with this distribution
338 Used by the `ping' plugin to send and receive ICMP packets.
341 Used to capture packets by the `dns' plugin.
344 Obviously used by the `perl' plugin. The library has to be compiled with
345 ithread support (introduced in Perl 5.6.0).
347 * librrd (optional; headers and library; rrdtool 1.0 and 1.2 both work fine)
348 If built without `librrd' the resulting binary will be `client only', i.e.
349 will send its values via multicast and not create any RRD files itself.
350 Alternatively you can chose to write CSV-files (Comma Separated Values)
353 * libsensors (optional)
354 To read from `lm_sensors', see the `sensors' plugin.
356 * libstatgrab may be used to collect statistics on systems other than Linux
357 and/or Solaris. Note that CPU- and disk-statistics, while being provided
358 by this library, are not supported in collectd right now..
359 <http://www.i-scream.org/libstatgrab/>
361 * libupsclient/nut (optional)
362 For the `nut' plugin which queries nut's `upsd'.
366 * librt, libsocket, libkstat, libdevinfo
367 Various standard Solaris libraries which provide system functions.
369 * CoreFoundation.framework and IOKit.framework
370 For compiling on Darwin in general and the `apple_sensors' plugin in
374 Collect statistics from virtual machines.
377 Parse XML data. This is needed for the `ascent' and `libvirt' plugins.
380 Configuring / Compiling / Installing
381 ------------------------------------
383 To configure, build and install collectd with the default settings, run
384 `./configure && make && make install'. For detailed, generic instructions
385 see INSTALL. For a complete list of configure options and their description,
386 run `./configure --help'.
388 By default, the configure script will check for all build dependencies and
389 disable all plugins whose requirements cannot be fulfilled (any other plugin
390 will be enabled). To enable a plugin, install missing dependencies (see
391 section `Prerequisites' above) and rerun `configure'. If you specify the
392 `--enable-<plugin>' configure option, you can force the plugin to be built.
393 This will most likely fail though unless you're working in a very unusual
394 setup and you really know what you're doing.
396 By default, collectd will be installed into `/opt/collectd'. You can adjust
397 this setting by specifying the `--prefix' configure option - see INSTALL for
398 details. If you pass DESTDIR=<path> to `make install', <path> will be
399 prefixed to all installation directories. This might be useful when creating
400 packages for collectd.
406 To compile correctly collectd needs to be able to initialize static
407 variables to NAN (Not A Number). Some C libraries, especially the GNU
408 libc, have a problem with that.
410 Luckily, with GCC it's possible to work around that problem: One can define
411 NAN as being (0.0 / 0.0) and `isnan' as `f != f'. However, to test this
412 ``implementation'' the configure script needs to compile and run a short
413 test program. Obviously running a test program when doing a cross-
414 compilation is, well, challenging.
416 If you run into this problem, you can use the `--with-nan-emulation'
417 configure option to force the use of this implementation. We can't promise
418 that the compiled binary actually behaves as it should, but since NANs
419 are likely never passed to the libm you have a good chance to be lucky.
425 For questions, bugreports, development information and basically all other
426 concerns please send an email to collectd's mailinglist at
427 <collectd at verplant.org>.
429 For live discussion and more personal contact visit us in IRC, we're in
430 channel #collectd on freenode.
436 Florian octo Forster <octo at verplant.org>,
437 Sebastian tokkee Harl <sh at tokkee.org>,
438 and many contributors (see `AUTHORS').
440 Please send bugreports and patches to the mailinglist, see `Contact' above.