1RRDTOOL(1) rrdtool RRDTOOL(1)
2
6 rrdtool - Round Robin Database Tool
7
9 rrdtool - [workdir]| function
10
12 OVERVIEW
13 It is pretty easy to gather status information from all sorts of
14 things, ranging from the temperature in your office to the number of
15 octets which have passed through the FDDI interface of your router. But
16 it is not so trivial to store this data in an efficient and systematic
17 manner. This is where RRDtool comes in handy. It lets you log and
18 analyze the data you gather from all kinds of data-sources (DS). The
19 data analysis part of RRDtool is based on the ability to quickly
20 generate graphical representations of the data values collected over a
21 definable time period.
22 In this man page you will find general information on the design and
24 functionality of the Round Robin Database Tool (RRDtool). For a more
25 detailed description of how to use the individual functions of RRDtool
26 check the corresponding man page.
27 For an introduction to the usage of RRDtool make sure you consult the
29 rrdtutorial.
30 FUNCTIONS
32 While the man pages talk of command line switches you have to set in
33 order to make RRDtool work it is important to note that RRDtool can be
34 remotely controlled through a set of pipes. This saves a considerable
35 amount of startup time when you plan to make RRDtool do a lot of things
36 quickly. Check the section on Remote_Control further down. There is
37 also a number of language bindings for RRDtool which allow you to use
38 it directly from Perl, python, Tcl, PHP, etc.
39 create Set up a new Round Robin Database (RRD). Check rrdcreate.
41 update Store new data values into an RRD. Check rrdupdate.
43 updatev Operationally equivalent to update except for output. Check
45 rrdupdate.
46 graph Create a graph from data stored in one or several RRDs. Apart
48 from generating graphs, data can also be extracted to stdout.
49 Check rrdgraph.
50 dump Dump the contents of an RRD in plain ASCII. In connection with
52 restore you can use this to move an RRD from one computer
53 architecture to another. Check rrddump.
54 restore Restore an RRD in XML format to a binary RRD. Check rrdrestore
56 fetch Get data for a certain time period from a RRD. The graph
58 function uses fetch to retrieve its data from an RRD. Check
59 rrdfetch.
60 tune Alter setup of an RRD. Check rrdtune.
62 last Find the last update time of an RRD. Check rrdlast.
64 info Get information about an RRD. Check rrdinfo.
66 rrdresize
68 Change the size of individual RRAs. This is dangerous! Check
69 rrdresize.
70 xport Export data retrieved from one or several RRDs. Check rrdxport.
72 flushcached
74 Flush the values for a specific RRD file from memory. Check
75 rrdflushcached.
76 rrdcgi This is a standalone tool for producing RRD graphs on the fly.
78 Check rrdcgi.
79 HOW DOES RRDTOOL WORK?
81 Data Acquisition
82 When monitoring the state of a system, it is convenient to have
83 the data available at a constant time interval. Unfortunately,
84 you may not always be able to fetch data at exactly the time
85 you want to. Therefore RRDtool lets you update the log file at
86 any time you want. It will automatically interpolate the value
87 of the data-source (DS) at the latest official time-slot
88 (interval) and write this interpolated value to the log. The
89 original value you have supplied is stored as well and is also
90 taken into account when interpolating the next log entry.
91 Consolidation
93 You may log data at a 1 minute interval, but you might also be
94 interested to know the development of the data over the last
95 year. You could do this by simply storing the data in 1 minute
96 intervals for the whole year. While this would take
97 considerable disk space it would also take a lot of time to
98 analyze the data when you wanted to create a graph covering the
99 whole year. RRDtool offers a solution to this problem through
100 its data consolidation feature. When setting up an Round Robin
101 Database (RRD), you can define at which interval this
102 consolidation should occur, and what consolidation function
103 (CF) (average, minimum, maximum, total, last) should be used to
104 build the consolidated values (see rrdcreate). You can define
105 any number of different consolidation setups within one RRD.
106 They will all be maintained on the fly when new data is loaded
107 into the RRD.
108 Round Robin Archives
110 Data values of the same consolidation setup are stored into
111 Round Robin Archives (RRA). This is a very efficient manner to
112 store data for a certain amount of time, while using a known
113 and constant amount of storage space.
114 It works like this: If you want to store 1'000 values in 5
116 minute interval, RRDtool will allocate space for 1'000 data
117 values and a header area. In the header it will store a pointer
118 telling which slots (value) in the storage area was last
119 written to. New values are written to the Round Robin Archive
120 in, you guessed it, a round robin manner. This automatically
121 limits the history to the last 1'000 values (in our example).
122 Because you can define several RRAs within a single RRD, you
123 can setup another one, for storing 750 data values at a 2 hour
124 interval, for example, and thus keep a log for the last two
125 months at a lower resolution.
126 The use of RRAs guarantees that the RRD does not grow over time
128 and that old data is automatically eliminated. By using the
129 consolidation feature, you can still keep data for a very long
130 time, while gradually reducing the resolution of the data along
131 the time axis.
132 Using different consolidation functions (CF) allows you to
134 store exactly the type of information that actually interests
135 you: the maximum one minute traffic on the LAN, the minimum
136 temperature of your wine cellar, the total minutes of down
137 time, etc.
138 Unknown Data
140 As mentioned earlier, the RRD stores data at a constant
141 interval. Sometimes it may happen that no new data is available
142 when a value has to be written to the RRD. Data acquisition may
143 not be possible for one reason or other. With RRDtool you can
144 handle these situations by storing an *UNKNOWN* value into the
145 database. The value '*UNKNOWN*' is supported through all the
146 functions of the tool. When consolidating a data set, the
147 amount of *UNKNOWN* data values is accounted for and when a new
148 consolidated value is ready to be written to its Round Robin
149 Archive (RRA), a validity check is performed to make sure that
150 the percentage of unknown values in the data point is above a
151 configurable level. If not, an *UNKNOWN* value will be written
152 to the RRA.
153 Graphing
155 RRDtool allows you to generate reports in numerical and
156 graphical form based on the data stored in one or several RRDs.
157 The graphing feature is fully configurable. Size, color and
158 contents of the graph can be defined freely. Check rrdgraph for
159 more information on this.
160 Aberrant Behavior Detection
162 by Jake Brutlag
163 RRDtool provides the building blocks for near real-time
165 aberrant behavior detection. These components include:
166 · An algorithm for predicting the value of a time series one
168 time step into the future.
169 · A measure of deviation between predicted and observed
171 values.
172 · A mechanism to decide if and when an observed value or
174 sequence of observed values is too deviant from the
175 predicted value(s).
176 Here is a brief explanation of these components:
178 The Holt-Winters time series forecasting algorithm is an on-
180 line (or incremental) algorithm that adaptively predicts future
181 observations in a time series. Its forecast is the sum of three
182 components: a baseline (or intercept), a linear trend over time
183 (or slope), and a seasonal coefficient (a periodic effect, such
184 as a daily cycle). There is one seasonal coefficient for each
185 time point in the period (cycle). After a value is observed,
186 each of these components is updated via exponential smoothing.
187 This means that the algorithm "learns" from past values and
188 uses them to predict the future. The rate of adaptation is
189 governed by 3 parameters, alpha (intercept), beta (slope), and
190 gamma (seasonal). The prediction can also be viewed as a
191 smoothed value for the time series.
192 The measure of deviation is a seasonal weighted absolute
194 deviation. The term seasonal means deviation is measured
195 separately for each time point in the seasonal cycle. As with
196 Holt-Winters forecasting, deviation is predicted using the
197 measure computed from past values (but only at that point in
198 the seasonal cycle). After the value is observed, the algorithm
199 learns from the observed value via exponential smoothing.
200 Confidence bands for the observed time series are generated by
201 scaling the sequence of predicted deviation values (we usually
202 think of the sequence as a continuous line rather than a set of
203 discrete points).
204 Aberrant behavior (a potential failure) is reported whenever
206 the number of times the observed value violates the confidence
207 bands meets or exceeds a specified threshold within a specified
208 temporal window (e.g. 5 violations during the past 45 minutes
209 with a value observed every 5 minutes).
210 This functionality is embedded in a set of related RRAs. In
212 particular, a FAILURES RRA logs potential failures. With these
213 data you could, for example, use a front-end application to
214 RRDtool to initiate real-time alerts.
215 For a detailed description on how to set this up, see
217 rrdcreate.
218 REMOTE CONTROL
220 When you start RRDtool with the command line option '-' it waits for
221 input via standard input (STDIN). With this feature you can improve
222 performance by attaching RRDtool to another process (MRTG is one
223 example) through a set of pipes. Over these pipes RRDtool accepts the
224 same arguments as on the command line and some special commands like
225 quit, cd, mkdir and ls. For detailed help on the server commands type:
226 rrdtool help cd|mkdir|pwd|ls|quit
228 When a command is completed, RRDtool will print the string '"OK"',
230 followed by timing information of the form u:usertime s:systemtime.
231 Both values are the running totals of seconds since RRDtool was
232 started. If an error occurs, a line of the form '"ERROR:" Description
233 of error' will be printed instead. RRDtool will not abort, unless
234 something really serious happens. If a workdir is specified and the UID
235 is 0, RRDtool will do a chroot to that workdir. If the UID is not 0,
236 RRDtool only changes the current directory to workdir.
237 RRD Server
239 If you want to create a RRD-Server, you must choose a TCP/IP Service
240 number and add them to /etc/services like this:
241 rrdsrv 13900/tcp # RRD server
243 Attention: the TCP port 13900 isn't officially registered for rrdsrv.
245 You can use any unused port in your services file, but the server and
246 the client system must use the same port, of course.
247 With this configuration you can add RRDtool as meta-server to
249 /etc/inetd.conf. For example:
250 rrdsrv stream tcp nowait root /opt/rrd/bin/rrdtool rrdtool - /var/rrd
252 Don't forget to create the database directory /var/rrd and reinitialize
254 your inetd.
255 If all was setup correctly, you can access the server with Perl
257 sockets, tools like netcat, or in a quick interactive test by using
258 'telnet localhost rrdsrv'.
259 NOTE: that there is no authentication with this feature! Do not setup
261 such a port unless you are sure what you are doing.
262
264 For very big setups, updating thousands of RRD files often becomes a
265 serious IO problem. If you run into such problems, you might want to
266 take a look at rrdcached, a caching daemon for RRDtool which may help
267 you lessen the stress on your disks.
268
270 rrdcreate, rrdupdate, rrdgraph, rrddump, rrdfetch, rrdtune, rrdlast,
271 rrdxport, rrdflushcached, rrdcached
272
274 Bugs? Features!
275
277 Tobias Oetiker <tobi@oetiker.ch>
2781.4.4 2009-10-14 RRDTOOL(1)