1 collectd - System information collection daemon
2 =================================================
8 collectd is a small daemon which collects statistics about a computer's
9 usage and writes then into RRD files.
15 * collectd is able to collect the following data:
18 Apache server utilization: Number of bytes transfered, number of
19 requests handled and detailed scoreboard statistics
22 APC UPS Daemon: UPS charge, load, input/output/battery voltage, etc.
25 Sensors in Macs running Mac OS X / Darwin: Temperature, fanspeed and
29 Batterycharge, -current and voltage of ACPI and PMU based laptop
33 CPU utilization: Time spent in the system, user, nice, idle, and related
37 CPU frequency (For laptops with speed step or a similar technology)
40 Mountpoint usage (Basically the values `df(1)' delivers)
43 Disk utilization: Sectors read/written, number of read/write actions,
44 average time an IO-operation took to complete.
47 DNS traffic: Query types, response codes, opcodes and traffic/octets
51 Email statistics: Count, traffic, spam scores and checks.
52 See collectd-email(5).
55 Amount of entropy available to the system.
58 Values gathered by a custom program or script.
62 Harddisk temperatures using hddtempd.
65 Interface traffic: Number of octets, packets and errors for each
69 Iptables' counters: Number of bytes that were matched by a certain
73 IPVS connection statistics (number of connections, octets and packets
74 for each service and destination).
75 See http://www.linuxvirtualserver.org/software/index.html.
78 IRQ counters: Frequency in which certain interrupts occur.
81 System load average over the last 1, 5 and 15 minutes.
84 Motherboard sensors: temperature, fanspeed and voltage information,
88 Statistics of the memcached distributed caching system.
89 <http://www.danga.com/memcached/>
92 Memory utilization: Memory occupied by running processes, page cache,
93 buffer cache and free.
96 Information provided by serial multimeters, such as the `Metex
100 MySQL server statistics: Commands issued, handlers triggered, thread
101 usage, query cache utilization and traffic/octets sent and received.
104 Very detailed Linux network interface and routing statistics. You can get
105 (detailed) information on interfaces, qdiscs, classes, and, if you can
106 make use of it, filters.
109 Receive values that were collected by other hosts. Large setups will
110 want to collect the data on one dedicated machine, and this is the
111 plugin of choice for that.
114 NFS Procedures: Which NFS command were called how often. Only NFSv2 and
118 Collects statistics from `nginx' (speak: engine X), a HTTP and mail
122 NTP daemon statistics: Local clock drift, offset to peers, etc.
125 Network UPS tools: UPS current, voltage, power, charge, utilisation,
126 temperature, etc. See upsd(8).
129 The perl plugin implements a Perl-interpreter into collectd. You can
130 write your own plugins in Perl and return arbitrary values using this
131 API. See collectd-perl(5).
133 This plugin is still considered to be experimental and subject to change
134 between minor releases.
137 Network latency: Time to reach the default gateway or another given
141 Process counts: Number of running, sleeping, zombie, ... processes.
144 System sensors, accessed using lm_sensors: Voltages, temperatures and
148 RX and TX of serial interfaces. Linux only; needs root privileges.
151 Read values from SNMP (Simple Network Management Protocol) enabled
152 network devices such as switches, routers, thermometers, rack monitoring
153 servers, etc. See collectd-snmp(5).
156 Pages swapped out onto harddisk or whatever is called `swap' by the OS..
159 Bytes and operations read and written on tape devices. Solaris only.
162 Number of TCP connections to specific local and remote ports.
165 Users currently logged in.
168 System resources used by Linux VServers.
169 See <http://linux-vserver.org/>.
172 Link quality of wireless cards. Linux only.
175 Bitrate and frequency of music played with XMMS.
177 * Output can be written or send to various destinations by the following
181 Write to comma separated values (CSV) files. This needs lots of
182 diskspace but is extremely portable and can be analysed with almost
183 every program that can analyse anything. Even Microsoft's Excel..
186 Send the data to a remote host to save the data somehow. This is useful
187 for large setups where the data should be saved by a dedicated machine.
190 Of course the values are propagated to plugins written in Perl, too, so
191 you can easily do weird stuff with the plugins we didn't dare think of
192 ;) See collectd-perl(5).
195 Output to round-robin-database (RRD) files using librrd. See rrdtool(1).
196 This is likely the most popular destination for such values. Since
197 updates to RRD-files are somewhat expensive this plugin can cache
198 updates to the files and write a bunch of updates at once, which lessens
202 One can query the values from the unixsock plugin whenever they're
203 needed. Please read collectd-unixsock(5) for a description on how that's
206 * Logging is, as everything in collectd, provided by plugins. The following
207 plugins keep up informed about what's going on:
210 Writes logmessages to a file or STDOUT/STDERR.
213 Logs to the standard UNIX logging mechanism, syslog.
215 * Performance: Since collectd is running as a daemon it doesn't spend much
216 time starting up again and again. With the exception of the exec plugin no
217 processes are forked. Caching in output plugins, such as the rrdtool and
218 network plugins, makes sure your resources are used efficiently. Also,
219 since collectd is programmed multithreaded it benefits from hyperthreading
220 and multicore processors and makes sure that the daemon isn't idle if only
221 one plugins waits for an IO-operation to complete.
223 * Once set up, hardly any maintenance is necessary. Setup is kept as easy
224 as possible and the default values should be okay for most users.
230 * collectd's configuration file can be found at `sysconfdir'/collectd.conf.
231 Run `collectd -h' for a list of builtin defaults. See `collectd.conf(5)'
232 for a list of options and a syntax description.
234 * When the `csv' or `rrdtool' plugins are loaded they'll write the values to
235 files. The usual place for these files is beneath `/var/lib/collectd'.
237 * When using some of the plugins, collectd needs to run as user root, since
238 only root can do certain things, such as craft ICMP packages needed to ping
239 other hosts. collectd should NOT be installed setuid root since it can be
240 used to overwrite valuable files!
242 * Sample scripts to generate graphs reside in `contrib/' in the source
243 package or somewhere near `/usr/share/doc/collectd' in most distributions.
244 Please be aware that those script are meant as a starting point for your
245 own experiments.. Some of them require the `RRDs' Perl module.
246 (`librrds-perl' on Debian) If you have written a more sophisticated
247 solution please share it with us.
249 * The RRAs of the automatically created RRD files depend on the `step'
250 and `heartbeat' settings given. If change these settings you may need to
251 re-create the files, losing all data. Please be aware of that when changing
252 the values and read the rrdtool(1) manpage thoroughly.
255 collectd and chkrootkit
256 -----------------------
258 If you are using the `dns' plugin chkrootkit(1) will report collectd as a
259 packet sniffer ("<iface>: PACKET SNIFFER(/usr/sbin/collectd[<pid>])"). The
260 plugin captures all UDP packets on port 53 to analyze the DNS traffic. In
261 this case, collectd is a legitimate sniffer and the report should be
262 considered to be a false positive. However, you might want to check that
263 this really is collectd and not some other, illegitimate sniffer.
269 To compile collectd from source you will need:
271 * Usual suspects: C compiler, linker, preprocessor, make, ...
273 * A POSIX-threads (pthread) implementation.
274 Since gathering some statistics is slow (network connections, slow devices,
275 etc) the collectd is parallelized. The POSIX threads interface is being
276 used and should be found in various implementations for hopefully all
280 If you want to use the `apache' and/or `nginx' plugins.
283 For querying iptables counters.
285 * libmysqlclient (optional)
286 Unsurprisingly used by the `mysql' plugin.
288 * libnetlink (optional)
289 Used, obviously, for the `netlink' plugin.
291 * libnetsnmp (optional)
292 For the `snmp' plugin.
294 * liboping (optional, if not found a version shipped with this distribution
296 Used by the `ping' plugin to send and receive ICMP packets.
299 Used to capture packets by the `dns' plugin.
301 * librrd (optional; headers and library; rrdtool 1.0 and 1.2 both work fine)
302 If built without `librrd' the resulting binary will be `client only', i.e.
303 will send its values via multicast and not create any RRD files itself.
304 Alternatively you can chose to write CSV-files (Comma Separated Values)
307 * libsensors (optional)
308 To read from `lm_sensors', see the `sensors' plugin.
310 * libstatgrab may be used to collect statistics on systems other than Linux
311 and/or Solaris. Note that CPU- and disk-statistics, while being provided
312 by this library, are not supported in collectd right now..
313 <http://www.i-scream.org/libstatgrab/>
315 * libupsclient/nut (optional)
316 For the `nut' plugin which queries nut's `upsd'.
320 * librt, libsocket, libkstat, libdevinfo
321 Various standard Solaris libraries which provide system functions.
323 * CoreFoundation.framework and IOKit.framework
324 For compiling on Darwin in general and the `apple_sensors' plugin in
328 Configuring / Compiling / Installing
329 ------------------------------------
331 To configure, build and install collectd with the default settings, run
332 `./configure && make && make install'. For detailed, generic instructions
333 see INSTALL. For a complete list of configure options and their description,
334 run `./configure --help'.
336 By default, the configure script will check for all build dependencies and
337 disable all plugins whose requirements cannot be fulfilled (any other plugin
338 will be enabled). To enable a plugin, install missing dependencies (see
339 section `Prerequisites' above) and rerun `configure'. If you specify the
340 `--enable-<plugin>' configure option, you can force the plugin to be built.
341 This will most likely fail though unless you're working in a very unusual
342 setup and you really know what you're doing.
344 By default, collectd will be installed into `/opt/collectd'. You can adjust
345 this setting by specifying the `--prefix' configure option - see INSTALL for
346 details. If you pass DESTDIR=<path> to `make install', <path> will be
347 prefixed to all installation directories. This might be useful when creating
348 packages for collectd.
354 To compile correctly collectd needs to be able to initialize static
355 variables to NAN (Not A Number). Some C libraries, especially the GNU
356 libc, have a problem with that.
358 Luckily, with GCC it's possible to work around that problem: One can define
359 NAN as being (0.0 / 0.0) and `isnan' as `f != f'. However, to test this
360 ``implementation'' the configure script needs to compile and run a short
361 test program. Obviously running a test program when doing a cross-
362 compilation is, well, challenging.
364 If you run into this problem, you can use the `--with-nan-emulation'
365 configure option to force the use of this implementation. We can't promise
366 that the compiled binary actually behaves as it should, but since NANs
367 are likely never passed to the libm you have a good chance to be lucky.
373 For questions, bugreports, development information and basically all other
374 concerns please send an email to collectd's mailinglist at
375 <collectd at verplant.org>.
377 For live discussion and more personal contact visit us in IRC, we're in
378 channel #collectd on freenode.
384 Florian octo Forster <octo at verplant.org>,
385 Sebastian tokkee Harl <sh at tokkee.org>,
386 and many contributors (see `AUTHORS').
388 Please send bugreports and patches to the mailinglist, see `Contact' above.