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 Various sensors in the Aquaero 5 watercooling board made by Aquacomputer.
33 Statistics about Ascent, a free server for the game `World of Warcraft'.
36 Batterycharge, -current and voltage of ACPI and PMU based laptop
40 Name server and resolver statistics from the `statistics-channel'
41 interface of BIND 9.5, 9,6 and later.
44 CPU accounting information for process groups under Linux.
47 Number of nf_conntrack entries.
50 Number of context switches done by the operating system.
53 CPU utilization: Time spent in the system, user, nice, idle, and related
57 CPU frequency (For laptops with speed step or a similar technology)
60 Parse statistics from websites using regular expressions.
63 Retrieves JSON data via cURL and parses it according to user
67 Retrieves XML data via cURL and parses it according to user
71 Executes SQL statements on various databases and interprets the returned
75 Mountpoint usage (Basically the values `df(1)' delivers)
78 Disk utilization: Sectors read/written, number of read/write actions,
79 average time an IO-operation took to complete.
82 DNS traffic: Query types, response codes, opcodes and traffic/octets
86 Email statistics: Count, traffic, spam scores and checks.
87 See collectd-email(5).
90 Amount of entropy available to the system.
93 Network interface card statistics.
96 Values gathered by a custom program or script.
100 Count the number of files in directories.
103 Linux file-system based caching framework statistics.
106 Receive multicast traffic from Ganglia instances.
109 Harddisk temperatures using hddtempd.
112 Interface traffic: Number of octets, packets and errors for each
116 Iptables' counters: Number of bytes that were matched by a certain
120 IPMI (Intelligent Platform Management Interface) sensors information.
123 IPVS connection statistics (number of connections, octets and packets
124 for each service and destination).
125 See http://www.linuxvirtualserver.org/software/index.html.
128 IRQ counters: Frequency in which certain interrupts occur.
131 Integrates a `Java Virtual Machine' (JVM) to execute plugins in Java
132 bytecode. See “Configuring with libjvm” below.
135 System load average over the last 1, 5 and 15 minutes.
138 Detailed CPU statistics of the “Logical Partitions” virtualization
139 technique built into IBM's POWER processors.
142 CPU, memory, disk and network I/O statistics from virtual machines.
145 Size of “Logical Volumes” (LV) and “Volume Groups” (VG) of Linux'
146 “Logical Volume Manager” (LVM).
149 Queries very detailed usage statistics from wireless LAN adapters and
150 interfaces that use the Atheros chipset and the MadWifi driver.
153 Motherboard sensors: temperature, fanspeed and voltage information,
157 Linux software-RAID device information (number of active, failed, spare
161 Query and parse data from a memcache daemon (memcached).
164 Statistics of the memcached distributed caching system.
165 <http://www.danga.com/memcached/>
168 Memory utilization: Memory occupied by running processes, page cache,
169 buffer cache and free.
172 Collects CPU usage, memory usage, temperatures and power consumption from
173 Intel Many Integrated Core (MIC) CPUs.
176 Reads values from Modbus/TCP enabled devices. Supports reading values
177 from multiple "slaves" so gateway devices can be used.
180 Information provided by serial multimeters, such as the `Metex
184 MySQL server statistics: Commands issued, handlers triggered, thread
185 usage, query cache utilization and traffic/octets sent and received.
188 Plugin to query performance values from a NetApp storage system using the
189 “Manage ONTAP” SDK provided by NetApp.
192 Very detailed Linux network interface and routing statistics. You can get
193 (detailed) information on interfaces, qdiscs, classes, and, if you can
194 make use of it, filters.
197 Receive values that were collected by other hosts. Large setups will
198 want to collect the data on one dedicated machine, and this is the
199 plugin of choice for that.
202 NFS Procedures: Which NFS command were called how often. Only NFSv2 and
206 Collects statistics from `nginx' (speak: engine X), a HTTP and mail
210 NTP daemon statistics: Local clock drift, offset to peers, etc.
213 Network UPS tools: UPS current, voltage, power, charge, utilisation,
214 temperature, etc. See upsd(8).
217 Information about Non-Uniform Memory Access (NUMA).
220 Queries routing information from the “Optimized Link State Routing”
223 - onewire (EXPERIMENTAL!)
224 Read onewire sensors using the owcapu library of the owfs project.
225 Please read in collectd.conf(5) why this plugin is experimental.
228 RX and TX of each client in openvpn-status.log (status-version 2).
229 <http://openvpn.net/index.php/documentation/howto.html>
232 Query data from an Oracle database.
235 The perl plugin implements a Perl-interpreter into collectd. You can
236 write your own plugins in Perl and return arbitrary values using this
237 API. See collectd-perl(5).
240 Query statistics from BSD's packet filter "pf".
243 Receive and dispatch timing values from Pinba, a profiling extension for
247 Network latency: Time to reach the default gateway or another given
251 PostgreSQL database statistics: active server connections, transaction
252 numbers, block IO, table row manipulations.
255 PowerDNS name server statistics.
258 Process counts: Number of running, sleeping, zombie, ... processes.
261 Counts various aspects of network protocols such as IP, TCP, UDP, etc.
264 The python plugin implements a Python interpreter into collectd. This
265 makes it possible to write plugins in Python which are executed by
266 collectd without the need to start a heavy interpreter every interval.
267 See collectd-python(5) for details.
270 The redis plugin gathers information from a redis server, including:
271 uptime, used memory, total connections etc.
274 Query interface and wireless registration statistics from RouterOS.
277 RRDtool caching daemon (RRDcacheD) statistics.
280 System sensors, accessed using lm_sensors: Voltages, temperatures and
284 RX and TX of serial interfaces. Linux only; needs root privileges.
287 Read values from SNMP (Simple Network Management Protocol) enabled
288 network devices such as switches, routers, thermometers, rack monitoring
289 servers, etc. See collectd-snmp(5).
292 Pages swapped out onto harddisk or whatever is called `swap' by the OS..
295 Parse table-like structured files.
298 Follows (tails) logfiles, parses them by lines and submits matched
302 Follows (tails) files in CSV format, parses each line and submits
306 Bytes and operations read and written on tape devices. Solaris only.
309 Number of TCP connections to specific local and remote ports.
312 TeamSpeak2 server statistics.
315 Plugin to read values from `The Energy Detective' (TED).
318 Linux ACPI thermal zone information.
321 Reads the number of records and file size from a running Tokyo Tyrant
325 System uptime statistics.
328 Users currently logged in.
331 Various statistics from Varnish, an HTTP accelerator.
334 Virtual memory statistics, e. g. the number of page-ins/-outs or the
335 number of pagefaults.
338 System resources used by Linux VServers.
339 See <http://linux-vserver.org/>.
342 Link quality of wireless cards. Linux only.
345 Bitrate and frequency of music played with XMMS.
348 Statistics for ZFS' “Adaptive Replacement Cache” (ARC).
350 * Output can be written or sent to various destinations by the following
354 Sends JSON-encoded data to an Advanced Message Queuing Protocol (AMQP)
355 server, such as RabbitMQ.
358 Write to comma separated values (CSV) files. This needs lots of
359 diskspace but is extremely portable and can be analysed with almost
360 every program that can analyse anything. Even Microsoft's Excel..
363 Send the data to a remote host to save the data somehow. This is useful
364 for large setups where the data should be saved by a dedicated machine.
367 Of course the values are propagated to plugins written in Perl, too, so
368 you can easily do weird stuff with the plugins we didn't dare think of
369 ;) See collectd-perl(5).
372 It's possible to implement write plugins in Python using the python
373 plugin. See collectd-python(5) for details.
376 Output to round-robin-database (RRD) files using the RRDtool caching
377 daemon (RRDcacheD) - see rrdcached(1). That daemon provides a general
378 implementation of the caching done by the `rrdtool' plugin.
381 Output to round-robin-database (RRD) files using librrd. See rrdtool(1).
382 This is likely the most popular destination for such values. Since
383 updates to RRD-files are somewhat expensive this plugin can cache
384 updates to the files and write a bunch of updates at once, which lessens
388 One can query the values from the unixsock plugin whenever they're
389 needed. Please read collectd-unixsock(5) for a description on how that's
393 Sends data to Carbon, the storage layer of Graphite using TCP or UDP. It
394 can be configured to avoid logging send errors (especially useful when
398 Sends the values collected by collectd to a web-server using HTTP POST
399 requests. The transmitted data is either in a form understood by the
400 Exec plugin or formatted in JSON.
403 Sends data to MongoDB, a NoSQL database.
406 Sends the values to a Redis key-value database server.
409 Sends data to Riemann, a stream processing and monitoring system.
411 * Logging is, as everything in collectd, provided by plugins. The following
412 plugins keep up informed about what's going on:
415 Writes logmessages to a file or STDOUT/STDERR.
418 Log messages are propagated to plugins written in Perl as well.
419 See collectd-perl(5).
422 It's possible to implement log plugins in Python using the python plugin.
423 See collectd-python(5) for details.
426 Logs to the standard UNIX logging mechanism, syslog.
428 * Notifications can be handled by the following plugins:
431 Send a desktop notification to a notification daemon, as defined in
432 the Desktop Notification Specification. To actually display the
433 notifications, notification-daemon is required.
434 See http://www.galago-project.org/specs/notification/.
437 Send an E-mail with the notification message to the configured
441 Execute a program or script to handle the notification.
442 See collectd-exec(5).
445 Writes the notification message to a file or STDOUT/STDERR.
448 Send the notification to a remote host to handle it somehow.
451 Notifications are propagated to plugins written in Perl as well.
452 See collectd-perl(5).
455 It's possible to implement notification plugins in Python using the
456 python plugin. See collectd-python(5) for details.
458 * Value processing can be controlled using the "filter chain" infrastructure
459 and "matches" and "targets". The following plugins are available:
461 - match_empty_counter
462 Match counter values which are currently zero.
465 Match values using a hash function of the hostname.
468 Match values by their identifier based on regular expressions.
471 Match values with an invalid timestamp.
474 Select values by their data sources' values.
476 - target_notification
477 Create and dispatch a notification.
480 Replace parts of an identifier using regular expressions.
483 Scale (multiply) values by an arbitrary value.
486 Set (overwrite) entire parts of an identifier.
488 * Miscellaneous plugins:
491 Selects multiple value lists based on patterns or regular expressions
492 and creates new aggregated values lists from those.
495 Checks values against configured thresholds and creates notifications if
496 values are out of bounds. See collectd-threshold(5) for details.
499 Sets the hostname to an unique identifier. This is meant for setups
500 where each client may migrate to another physical host, possibly going
501 through one or more name changes in the process.
503 * Performance: Since collectd is running as a daemon it doesn't spend much
504 time starting up again and again. With the exception of the exec plugin no
505 processes are forked. Caching in output plugins, such as the rrdtool and
506 network plugins, makes sure your resources are used efficiently. Also,
507 since collectd is programmed multithreaded it benefits from hyperthreading
508 and multicore processors and makes sure that the daemon isn't idle if only
509 one plugin waits for an IO-operation to complete.
511 * Once set up, hardly any maintenance is necessary. Setup is kept as easy
512 as possible and the default values should be okay for most users.
518 * collectd's configuration file can be found at `sysconfdir'/collectd.conf.
519 Run `collectd -h' for a list of builtin defaults. See `collectd.conf(5)'
520 for a list of options and a syntax description.
522 * When the `csv' or `rrdtool' plugins are loaded they'll write the values to
523 files. The usual place for these files is beneath `/var/lib/collectd'.
525 * When using some of the plugins, collectd needs to run as user root, since
526 only root can do certain things, such as craft ICMP packages needed to ping
527 other hosts. collectd should NOT be installed setuid root since it can be
528 used to overwrite valuable files!
530 * Sample scripts to generate graphs reside in `contrib/' in the source
531 package or somewhere near `/usr/share/doc/collectd' in most distributions.
532 Please be aware that those script are meant as a starting point for your
533 own experiments.. Some of them require the `RRDs' Perl module.
534 (`librrds-perl' on Debian) If you have written a more sophisticated
535 solution please share it with us.
537 * The RRAs of the automatically created RRD files depend on the `step'
538 and `heartbeat' settings given. If change these settings you may need to
539 re-create the files, losing all data. Please be aware of that when changing
540 the values and read the rrdtool(1) manpage thoroughly.
543 collectd and chkrootkit
544 -----------------------
546 If you are using the `dns' plugin chkrootkit(1) will report collectd as a
547 packet sniffer ("<iface>: PACKET SNIFFER(/usr/sbin/collectd[<pid>])"). The
548 plugin captures all UDP packets on port 53 to analyze the DNS traffic. In
549 this case, collectd is a legitimate sniffer and the report should be
550 considered to be a false positive. However, you might want to check that
551 this really is collectd and not some other, illegitimate sniffer.
557 To compile collectd from source you will need:
559 * Usual suspects: C compiler, linker, preprocessor, make, ...
561 * A POSIX-threads (pthread) implementation.
562 Since gathering some statistics is slow (network connections, slow devices,
563 etc) the collectd is parallelized. The POSIX threads interface is being
564 used and should be found in various implementations for hopefully all
567 * CoreFoundation.framework and IOKit.framework (optional)
568 For compiling on Darwin in general and the `apple_sensors' plugin in
570 <http://developer.apple.com/corefoundation/>
572 * libclntsh (optional)
573 Used by the `oracle' plugin.
575 * libcredis (optional)
576 Used by the redis plugin. Please note that you require a 0.2.2 version
577 or higher. <http://code.google.com/p/credis/>
580 If you want to use the `apache', `ascent', `curl', `nginx', or `write_http'
582 <http://curl.haxx.se/>
585 Used by the `dbi' plugin to connect to various databases.
586 <http://libdbi.sourceforge.net/>
588 * libesmtp (optional)
589 For the `notify_email' plugin.
590 <http://www.stafford.uklinux.net/libesmtp/>
592 * libganglia (optional)
593 Used by the `gmond' plugin to process data received from Ganglia.
594 <http://ganglia.info/>
596 * libgcrypt (optional)
597 Used by the `network' plugin for encryption and authentication.
598 <http://www.gnupg.org/>
601 If present, the uuid plugin will check for UUID from HAL.
602 <http://hal.freedesktop.org/>
605 For querying iptables counters.
606 <http://netfilter.org/>
608 If not found on the system, a version shipped with this distribution can
609 be used. It requires some Linux headers in /usr/include/linux. You can
610 force the build system to use the shipped version by specifying
611 --with-libiptc=shipped
612 when running the configure script.
615 Library that encapsulates the `Java Virtual Machine' (JVM). This library is
616 used by the Java plugin to execute Java bytecode. See “Configuring with
618 <http://openjdk.java.net/> (and others)
620 * libmemcached (optional)
621 Used by the `memcachec' plugin to connect to a memcache daemon.
622 <http://tangent.org/552/libmemcached.html>
624 * libmodbus (optional)
625 Used by the “modbus” plugin to communicate with Modbus/TCP devices. The
626 “modbus” plugin works with version 2.0.3 of the library – due to frequent
627 API changes other versions may or may not compile cleanly.
628 <http://www.libmodbus.org/>
630 * libmysqlclient (optional)
631 Unsurprisingly used by the `mysql' plugin.
632 <http://dev.mysql.com/>
634 * libnetapp (optional)
635 Required for the “netapp” plugin.
636 This library is part of the “Manage ONTAP SDK” published by NetApp.
638 * libnetlink (optional)
639 Used, obviously, for the `netlink' plugin.
640 <http://www.linuxfoundation.org/en/Net:Iproute2>
642 * libnetsnmp (optional)
643 For the `snmp' plugin.
644 <http://www.net-snmp.org/>
646 * libnotify (optional)
647 For the `notify_desktop' plugin.
648 <http://www.galago-project.org/>
650 * liboping (optional)
651 Used by the `ping' plugin to send and receive ICMP packets.
652 <http://verplant.org/liboping/>
654 * libowcapi (optional)
655 Used by the `onewire' plugin to read values from onewire sensors (or the
657 <http://www.owfs.org/>
660 Used to capture packets by the `dns' plugin.
661 <http://www.tcpdump.org/>
663 * libperfstat (optional)
664 Used by various plugins to gather statistics under AIX.
667 Obviously used by the `perl' plugin. The library has to be compiled with
668 ithread support (introduced in Perl 5.6.0).
669 <http://www.perl.org/>
672 The PostgreSQL C client library used by the `postgresql' plugin.
673 <http://www.postgresql.org/>
675 * libprotobuf-c, protoc-c (optional)
676 Used by the `pinba' plugin to generate a parser for the network packets
677 sent by the Pinba PHP extension.
678 <http://code.google.com/p/protobuf-c/>
680 * libpython (optional)
681 Used by the `python' plugin. Currently, Python 2.3 and later and Python 3
683 <http://www.python.org/>
685 * librabbitmq (optional; also called “rabbitmq-c”)
686 Used by the AMQP plugin for AMQP connections, for example to RabbitMQ.
687 <http://hg.rabbitmq.com/rabbitmq-c/>
689 * librouteros (optional)
690 Used by the `routeros' plugin to connect to a device running `RouterOS'.
691 <http://verplant.org/librouteros/>
694 Used by the `rrdtool' and `rrdcached' plugins. The latter requires RRDtool
695 client support which was added after version 1.3 of RRDtool. Versions 1.0,
696 1.2 and 1.3 are known to work with the `rrdtool' plugin.
697 <http://oss.oetiker.ch/rrdtool/>
699 * librt, libsocket, libkstat, libdevinfo (optional)
700 Various standard Solaris libraries which provide system functions.
701 <http://developers.sun.com/solaris/>
703 * libsensors (optional)
704 To read from `lm_sensors', see the `sensors' plugin.
705 <http://www.lm-sensors.org/>
707 * libstatgrab (optional)
708 Used by various plugins to collect statistics on systems other than Linux
710 <http://www.i-scream.org/libstatgrab/>
712 * libtokyotyrant (optional)
713 Used by the tokyotyrant plugin.
714 <http://1978th.net/tokyotyrant/>
716 * libupsclient/nut (optional)
717 For the `nut' plugin which queries nut's `upsd'.
718 <http://networkupstools.org/>
721 Collect statistics from virtual machines.
722 <http://libvirt.org/>
725 Parse XML data. This is needed for the `ascent' and `libvirt' plugins.
726 <http://xmlsoft.org/>
729 <http://www.xmms.org/>
732 Parse JSON data. This is needed for the `curl_json' plugin.
733 <http://github.com/lloyd/yajl>
735 * libvarnish (optional)
736 Fetches statistics from a Varnish instance. This is needed for the Varnish plugin
737 <http://varnish-cache.org>
739 Configuring / Compiling / Installing
740 ------------------------------------
742 To configure, build and install collectd with the default settings, run
743 `./configure && make && make install'. For detailed, generic instructions
744 see INSTALL. For a complete list of configure options and their description,
745 run `./configure --help'.
747 By default, the configure script will check for all build dependencies and
748 disable all plugins whose requirements cannot be fulfilled (any other plugin
749 will be enabled). To enable a plugin, install missing dependencies (see
750 section `Prerequisites' above) and rerun `configure'. If you specify the
751 `--enable-<plugin>' configure option, the script will fail if the depen-
752 dencies for the specified plugin are not met. In that case you can force the
753 plugin to be built using the `--enable-<plugin>=force' configure option.
754 This will most likely fail though unless you're working in a very unusual
755 setup and you really know what you're doing. If you specify the
756 `--disable-<plugin>' configure option, the plugin will not be built. If you
757 specify the `--enable-all-plugins' or `--disable-all-plugins' configure
758 options, all plugins will be enabled or disabled respectively by default.
759 Explicitly enabling or disabling a plugin overwrites the default for the
760 specified plugin. These options are meant for package maintainers and should
761 not be used in everyday situations.
763 By default, collectd will be installed into `/opt/collectd'. You can adjust
764 this setting by specifying the `--prefix' configure option - see INSTALL for
765 details. If you pass DESTDIR=<path> to `make install', <path> will be
766 prefixed to all installation directories. This might be useful when creating
767 packages for collectd.
769 Configuring with libjvm
770 -----------------------
772 To determine the location of the required files of a Java installation is not
773 an easy task, because the locations vary with your kernel (Linux, SunOS, …)
774 and with your architecture (x86, SPARC, …) and there is no ‘java-config’
775 script we could use. Configuration of the JVM library is therefore a bit
778 The easiest way to use the `--with-java=$JAVA_HOME' option, where
779 `$JAVA_HOME' is usually something like:
780 /usr/lib/jvm/java-1.5.0-sun-1.5.0.14
782 The configure script will then use find(1) to look for the following files:
788 If found, appropriate CPP-flags and LD-flags are set and the following
789 library checks succeed.
791 If this doesn't work for you, you have the possibility to specify CPP-flags,
792 C-flags and LD-flags for the ‘Java’ plugin by hand, using the following three
793 (environment) variables:
799 For example (shortened for demonstration purposes):
801 ./configure JAVA_CPPFLAGS="-I$JAVA_HOME/include -I$JAVA_HOME/include/linux"
803 Adding "-ljvm" to the JAVA_LDFLAGS is done automatically, you don't have to
809 To compile correctly collectd needs to be able to initialize static
810 variables to NAN (Not A Number). Some C libraries, especially the GNU
811 libc, have a problem with that.
813 Luckily, with GCC it's possible to work around that problem: One can define
814 NAN as being (0.0 / 0.0) and `isnan' as `f != f'. However, to test this
815 ``implementation'' the configure script needs to compile and run a short
816 test program. Obviously running a test program when doing a cross-
817 compilation is, well, challenging.
819 If you run into this problem, you can use the `--with-nan-emulation'
820 configure option to force the use of this implementation. We can't promise
821 that the compiled binary actually behaves as it should, but since NANs
822 are likely never passed to the libm you have a good chance to be lucky.
824 Likewise, collectd needs to know the layout of doubles in memory, in order
825 to craft uniform network packets over different architectures. For this, it
826 needs to know how to convert doubles into the memory layout used by x86. The
827 configure script tries to figure this out by compiling and running a few
828 small test programs. This is of course not possible when cross-compiling.
829 You can use the `--with-fp-layout' option to tell the configure script which
830 conversion method to assume. Valid arguments are:
832 * `nothing' (12345678 -> 12345678)
833 * `endianflip' (12345678 -> 87654321)
834 * `intswap' (12345678 -> 56781234)
840 For questions, bug reports, development information and basically all other
841 concerns please send an email to collectd's mailing list at
842 <collectd at verplant.org>.
844 For live discussion and more personal contact visit us in IRC, we're in
845 channel #collectd on freenode.
851 Florian octo Forster <octo at verplant.org>,
852 Sebastian tokkee Harl <sh at tokkee.org>,
853 and many contributors (see `AUTHORS').
855 Please send bug reports and patches to the mailing list, see `Contact'