1 RRDTOOL(1) rrdtool RRDTOOL(1)
6 rrdtool - Round Robin Database Tool
14 It is pretty easy to gather status information from all sorts of
15 things, ranging from the temperature in your office to the number of
16 octets which have passed through the FDDI interface of your router. But
17 it is not so trivial to store this data in an efficient and systematic
18 manner. This is where R\bRR\bRD\bDt\bto\boo\bol\bl comes in handy. It lets you _\bl_\bo_\bg _\ba_\bn_\bd _\ba_\bn_\ba_\b-
20 analysis part of RRDtool is based on the ability to quickly generate
21 graphical representations of the data values collected over a definable
22 time period.
24 In this man page you will find general information on the design and
25 functionality of the Round Robin Database Tool (RRDtool). For a more
27 check the corresponding man page.
29 For an introduction to the usage of RRDtool make sure you consult the
30 rrdtutorial.
34 While the man pages talk of command line switches you have to set in
35 order to make R\bRR\bRD\bDt\bto\boo\bol\bl work it is important to note that R\bRR\bRD\bDt\bto\boo\bol\bl can be
36 remotely controlled through a set of pipes. This saves a considerable
38 quickly. Check the section on Remote_Control further down. There is
39 also a number of language bindings for RRDtool which allow you to use
40 it directly from Perl, python, Tcl, PHP, etc.
46 u\bup\bpd\bda\bat\bte\bev\bv Operationally equivalent to u\bup\bpd\bda\bat\bte\be except for output. Check
47 rrdupdate.
50 from generating graphs, data can also be extracted to stdout.
51 Check rrdgraph.
54 restore you can use this to move an RRD from one computer
55 architecture to another. Check rrddump.
60 tion uses fetch to retrieve its data from an RRD. Check rrd-
61 fetch.
70 Change the size of individual RRAs. This is dangerous! Check
71 rrdresize.
76 Flush the values for a specific RRD file from memory. Check
77 rrdflushcached.
80 Check rrdcgi.
85 Data Acquisition
86 When monitoring the state of a system, it is convenient to have
87 the data available at a constant time interval. Unfortunately,
88 you may not always be able to fetch data at exactly the time
90 any time you want. It will automatically interpolate the value
92 (interval) and write this interpolated value to the log. The
93 original value you have supplied is stored as well and is also
94 taken into account when interpolating the next log entry.
96 Consolidation
97 You may log data at a 1 minute interval, but you might also be
98 interested to know the development of the data over the last
99 year. You could do this by simply storing the data in 1 minute
100 intervals for the whole year. While this would take consider-
101 able disk space it would also take a lot of time to analyze the
102 data when you wanted to create a graph covering the whole year.
104 solidation feature. When setting up an Round Robin Database
107 minimum, maximum, total, last) should be used to build the con-
108 solidated values (see rrdcreate). You can define any number of
112 Round Robin Archives
113 Data values of the same consolidation setup are stored into
115 store data for a certain amount of time, while using a known
116 and constant amount of storage space.
118 It works like this: If you want to store 1'000 values in 5
120 values and a header area. In the header it will store a pointer
121 telling which slots (value) in the storage area was last writ-
122 ten to. New values are written to the Round Robin Archive in,
123 you guessed it, a round robin manner. This automatically limits
124 the history to the last 1'000 values (in our example). Because
126 another one, for storing 750 data values at a 2 hour interval,
127 for example, and thus keep a log for the last two months at a
128 lower resolution.
131 and that old data is automatically eliminated. By using the
132 consolidation feature, you can still keep data for a very long
133 time, while gradually reducing the resolution of the data along
134 the time axis.
137 store exactly the type of information that actually interests
138 you: the maximum one minute traffic on the LAN, the minimum
139 temperature of your wine cellar, the total minutes of down
140 time, etc.
142 Unknown Data
144 val. Sometimes it may happen that no new data is available when
149 functions of the tool. When consolidating a data set, the
151 consolidated value is ready to be written to its Round Robin
153 the percentage of unknown values in the data point is above a
157 Graphing
160 graphing feature is fully configurable. Size, color and con-
161 tents of the graph can be defined freely. Check rrdgraph for
162 more information on this.
164 Aberrant Behavior Detection
165 by Jake Brutlag
168 rant behavior detection. These components include:
170 * An algorithm for predicting the value of a time series one
171 time step into the future.
173 * A measure of deviation between predicted and observed val-
174 ues.
176 * A mechanism to decide if and when an observed value or
178 dicted value(s).
180 Here is a brief explanation of these components:
182 The Holt-Winters time series forecasting algorithm is an on-
183 line (or incremental) algorithm that adaptively predicts future
184 observations in a time series. Its forecast is the sum of three
185 components: a baseline (or intercept), a linear trend over time
186 (or slope), and a seasonal coefficient (a periodic effect, such
187 as a daily cycle). There is one seasonal coefficient for each
188 time point in the period (cycle). After a value is observed,
189 each of these components is updated via exponential smoothing.
190 This means that the algorithm "learns" from past values and
191 uses them to predict the future. The rate of adaptation is gov-
192 erned by 3 parameters, alpha (intercept), beta (slope), and
193 gamma (seasonal). The prediction can also be viewed as a
194 smoothed value for the time series.
196 The measure of deviation is a seasonal weighted absolute devia-
198 for each time point in the seasonal cycle. As with Holt-Winters
199 forecasting, deviation is predicted using the measure computed
200 from past values (but only at that point in the seasonal
201 cycle). After the value is observed, the algorithm learns from
202 the observed value via exponential smoothing. Confidence bands
203 for the observed time series are generated by scaling the
204 sequence of predicted deviation values (we usually think of the
205 sequence as a continuous line rather than a set of discrete
206 points).
208 Aberrant behavior (a potential failure) is reported whenever
209 the number of times the observed value violates the confidence
210 bands meets or exceeds a specified threshold within a specified
211 temporal window (e.g. 5 violations during the past 45 minutes
212 with a value observed every 5 minutes).
216 data you could, for example, use a front-end application to
219 For a detailed description on how to set this up, see rrdcre-
220 ate.
225 input via standard input (STDIN). With this feature you can improve
231 rrdtool help cd|mkdir|pwd|ls|quit
233 When a command is completed, RRDtool will print the string '"OK"',
234 followed by timing information of the form u\bu:\b:_\bu_\bs_\be_\br_\bt_\bi_\bm_\be s\bs:\b:_\bs_\by_\bs_\bt_\be_\bm_\bt_\bi_\bm_\be.
235 Both values are the running totals of seconds since RRDtool was
236 started. If an error occurs, a line of the form '"ERROR:" _\bD_\be_\bs_\bc_\br_\bi_\bp_\bt_\bi_\bo_\bn
237 _\bo_\bf _\be_\br_\br_\bo_\br' will be printed instead. R\bRR\bRD\bDt\bto\boo\bol\bl will not abort, unless some-
239 0, RRDtool will do a chroot to that workdir. If the UID is not 0, RRD-
244 If you want to create a RRD-Server, you must choose a TCP/IP Service
247 rrdsrv 13900/tcp # RRD server
249 Attention: the TCP port 13900 isn't officially registered for rrdsrv.
250 You can use any unused port in your services file, but the server and
251 the client system must use the same port, of course.
253 With this configuration you can add RRDtool as meta-server to
256 rrdsrv stream tcp nowait root /opt/rrd/bin/rrdtool rrdtool - /var/rrd
258 Don't forget to create the database directory /var/rrd and reinitialize
259 your inetd.
261 If all was setup correctly, you can access the server with Perl sock-
262 ets, tools like netcat, or in a quick interactive test by using 'telnet
263 localhost rrdsrv'.
266 such a port unless you are sure what you are doing.
268 R\bRR\bRD\bDC\bCA\bAC\bCH\bHE\bED\bD,\b, T\bTH\bHE\bE C\bCA\bAC\bCH\bHI\bIN\bNG\bG D\bDA\bAE\bEM\bMO\bON\bN
269 For very big setups, updating thousands of RRD files often becomes a
270 serious IO problem. If you run into such problems, you might want to
271 take a look at rrdcached, a caching daemon for RRDTool which may help
272 you lessen the stress on your disks.
275 rrdcreate, rrdupdate, rrdgraph, rrddump, rrdfetch, rrdtune, rrdlast,
276 rrdxport, rrdflushcached, rrdcached
279 Bugs? Features!
282 Tobias Oetiker <tobi@oetiker.ch>
286 1.3.99909060808 2009-05-26 RRDTOOL(1)