1 RRDTUTORIAL(1) rrdtool RRDTUTORIAL(1)
6 rrdtutorial - Alex van den Bogaerdt's RRDtool tutorial
9 RRDtool is written by Tobias Oetiker <tobi@oetiker.ch> with
10 contributions from many people all around the world. This document is
11 written by Alex van den Bogaerdt <alex@vandenbogaerdt.nl> to help you
12 understand what RRDtool is and what it can do for you.
14 The documentation provided with RRDtool can be too technical for some
15 people. This tutorial is here to help you understand the basics of
16 RRDtool. It should prepare you to read the documentation yourself. It
17 also explains the general things about statistics with a focus on
18 networking.
22 Please don't skip ahead in this document! The first part of this
23 document explains the basics and may be boring. But if you don't
24 understand the basics, the examples will not be as meaningful to you.
26 Sometimes things change. This example used to provide numbers like
27 "0.04" in stead of "4.00000e-02". Those are really the same numbers,
28 just written down differently. Don't be alarmed if a future version of
29 rrdtool displays a slightly different form of output. The examples in
30 this document are correct for version 1.2.0 of RRDtool.
32 Also, sometimes bugs do occur. They may also influence the outcome of
33 the examples. Example speed4.png was suffering from this (the handling
34 of unknown data in an if-statement was wrong). Normal data will be just
35 fine (a bug in rrdtool wouldn't last long) but special cases like NaN,
36 INF and so on may last a bit longer. Try another version if you can,
37 or just live with it.
39 I fixed the speed4.png example (and added a note). There may be other
40 examples which suffer from the same or a similar bug. Try to fix it
41 yourself, which is a great excercise. But please do not submit your
42 result as a fix to the source of this document. Discuss it on the
43 user's list, or write to me.
46 RRDtool refers to Round Robin Database tool. Round robin is a
47 technique that works with a fixed amount of data, and a pointer to the
48 current element. Think of a circle with some dots plotted on the edge.
49 These dots are the places where data can be stored. Draw an arrow from
50 the center of the circle to one of the dots; this is the pointer. When
51 the current data is read or written, the pointer moves to the next
52 element. As we are on a circle there is neither a beginning nor an end,
53 you can go on and on and on. After a while, all the available places
54 will be used and the process automatically reuses old locations. This
55 way, the dataset will not grow in size and therefore requires no
56 maintenance. RRDtool works with with Round Robin Databases (RRDs). It
57 stores and retrieves data from them.
59 W\bWh\bha\bat\bt d\bda\bat\bta\ba c\bca\ban\bn b\bbe\be p\bpu\but\bt i\bin\bnt\bto\bo a\ban\bn R\bRR\bRD\bD?\b?
60 You name it, it will probably fit as long as it is some sort of time-
61 series data. This means you have to be able to measure some value at
62 several points in time and provide this information to RRDtool. If you
63 can do this, RRDtool will be able to store it. The values must be
64 numerical but don't have to be integers, as is the case with MRTG (the
65 next section will give more details on this more specialized
66 application).
68 Many examples below talk about SNMP which is an acronym for Simple
69 Network Management Protocol. "Simple" refers to the protocol. It does
70 not mean it is simple to manage or monitor a network. After working
71 your way through this document, you should know enough to be able to
72 understand what people are talking about. For now, just realize that
73 SNMP can be used to query devices for the values of counters they keep.
74 It is the value from those counters that we want to store in the RRD.
76 W\bWh\bha\bat\bt c\bca\ban\bn I\bI d\bdo\bo w\bwi\bit\bth\bh t\bth\bhi\bis\bs t\bto\boo\bol\bl?\b?
77 RRDtool originated from MRTG (Multi Router Traffic Grapher). MRTG
78 started as a tiny little script for graphing the use of a university's
79 connection to the Internet. MRTG was later (ab-)used as a tool for
80 graphing other data sources including temperature, speed, voltage,
81 number of printouts and the like.
83 Most likely you will start to use RRDtool to store and process data
84 collected via SNMP. The data will most likely be bytes (or bits)
85 transfered from and to a network or a computer. But it can also be
86 used to display tidal waves, solar radiation, power consumption, number
87 of visitors at an exhibition, noise levels near an airport, temperature
88 on your favorite holiday location, temperature in the fridge and
89 whatever you imagination can come up with.
91 You only need a sensor to measure the data and be able to feed the
92 numbers into RRDtool. RRDtool then lets you create a database, store
93 data in it, retrieve that data and create graphs in PNG format for
94 display on a web browser. Those PNG images are dependent on the data
95 you collected and could be, for instance, an overview of the average
96 network usage, or the peaks that occurred.
98 W\bWh\bha\bat\bt i\bif\bf I\bI s\bst\bti\bil\bll\bl h\bha\bav\bve\be p\bpr\bro\bob\bbl\ble\bem\bms\bs a\baf\bft\bte\ber\br r\bre\bea\bad\bdi\bin\bng\bg t\bth\bhi\bis\bs d\bdo\boc\bcu\bum\bme\ben\bnt\bt?\b?
99 First of all: read it again! You may have missed something. If you are
100 unable to compile the sources and you have a fairly common OS, it will
101 probably not be the fault of RRDtool. There may be pre-compiled
102 versions around on the Internet. If they come from trusted sources, get
103 one of those.
105 If on the other hand the program works but does not give you the
106 expected results, it will be a problem with configuring it. Review your
107 configuration and compare it with the examples that follow.
109 There is a mailing list and an archive of it. Read the list for a few
110 weeks and search the archive. It is considered rude to just ask a
111 question without searching the archives: your problem may already have
112 been solved for somebody else! This is true for most, if not all,
113 mailing lists and not only for this particular one. Look in the
114 documentation that came with RRDtool for the location and usage of the
115 list.
117 I suggest you take a moment to subscribe to the mailing list right now
118 by sending an email to <rrd-users-request@lists.oetiker.ch> with a
119 subject of "subscribe". If you ever want to leave this list, just write
120 an email to the same address but now with a subject of "unsubscribe".
123 By giving you some detailed descriptions with detailed examples. I
124 assume that following the instructions in the order presented will give
125 you enough knowledge of RRDtool to experiment for yourself. If it
126 doesn't work the first time, don't give up. Reread the stuff that you
127 did understand, you may have missed something.
129 By following the examples you get some hands-on experience and, even
130 more important, some background information of how it works.
132 You will need to know something about hexadecimal numbers. If you don't
133 then start with reading bin_dec_hex before you continue here.
135 Y\bYo\bou\bur\br f\bfi\bir\brs\bst\bt R\bRo\bou\bun\bnd\bd R\bRo\bob\bbi\bin\bn D\bDa\bat\bta\bab\bba\bas\bse\be
136 In my opinion the best way to learn something is to actually do it.
137 Why not start right now? We will create a database, put some values in
138 it and extract this data again. Your output should be the same as the
139 output that is included in this document.
141 We will start with some easy stuff and compare a car with a router, or
142 compare kilometers (miles if you wish) with bits and bytes. It's all
143 the same: some number over some time.
145 Assume we have a device that transfers bytes to and from the Internet.
146 This device keeps a counter that starts at zero when it is turned on,
147 increasing with every byte that is transfered. This counter will
148 probably have a maximum value. If this value is reached and an extra
149 byte is counted, the counter starts over at zero. This is the same as
150 many counters in the world such as the mileage counter in a car.
152 Most discussions about networking talk about bits per second so lets
153 get used to that right away. Assume a byte is eight bits and start to
154 think in bits not bytes. The counter, however, still counts bytes! In
155 the SNMP world most of the counters are 32 bits. That means they are
156 counting from 0 to 4294967295. We will use these values in the
157 examples. The device, when asked, returns the current value of the
158 counter. We know the time that has passes since we last asked so we now
159 know how many bytes have been transfered ***on average*** per second.
160 This is not very hard to calculate. First in words, then in
161 calculations:
163 1. Take the current counter, subtract the previous value from it.
165 2. Do the same with the current time and the previous time (in
166 seconds).
168 3. Divide the outcome of (1) by the outcome of (2), the result is the
169 amount of bytes per second. Multiply by eight to get the number of
170 bits per second (bps).
172 bps = (counter_now - counter_before) / (time_now - time_before) * 8
174 For some people it may help to translate this to an automobile example.
175 Do not try this example, and if you do, don't blame me for the results!
177 People who are not used to think in kilometers per hour can translate
178 most into miles per hour by dividing km by 1.6 (close enough). I will
179 use the following abbreviations:
181 m: meter
182 km: kilometer (= 1000 meters).
183 h: hour
184 s: second
185 km/h: kilometers per hour
186 m/s: meters per second
188 You are driving a car. At 12:05 you read the counter in the dashboard
189 and it tells you that the car has moved 12345 km until that moment. At
190 12:10 you look again, it reads 12357 km. This means you have traveled
191 12 km in five minutes. A scientist would translate that into meters per
192 second and this makes a nice comparison toward the problem of (bytes
193 per five minutes) versus (bits per second).
195 We traveled 12 kilometers which is 12000 meters. We did that in five
196 minutes or 300 seconds. Our speed is 12000m / 300s or 40 m/s.
198 We could also calculate the speed in km/h: 12 times 5 minutes is an
199 hour, so we have to multiply 12 km by 12 to get 144 km/h. For our
200 native English speaking friends: that's 90 mph so don't try this
201 example at home or where I live :)
203 Remember: these numbers are averages only. There is no way to figure
204 out from the numbers, if you drove at a constant speed. There is an
205 example later on in this tutorial that explains this.
207 I hope you understand that there is no difference in calculating m/s or
208 bps; only the way we collect the data is different. Even the k from
209 kilo is the same as in networking terms k also means 1000.
211 We will now create a database where we can keep all these interesting
212 numbers. The method used to start the program may differ slightly from
213 OS to OS, but I assume you can figure it out if it works different on
214 your's. Make sure you do not overwrite any file on your system when
215 executing the following command and type the whole line as one long
216 line (I had to split it for readability) and skip all of the '\'
217 characters.
219 rrdtool create test.rrd \
220 --start 920804400 \
221 DS:speed:COUNTER:600:U:U \
222 RRA:AVERAGE:0.5:1:24 \
223 RRA:AVERAGE:0.5:6:10
225 (So enter: "rrdtool create test.rrd --start 920804400 DS ...")
228 We created the round robin database called test (test.rrd) which starts
229 at noon the day I started writing this document, 7th of March, 1999
230 (this date translates to 920804400 seconds as explained below). Our
231 database holds one data source (DS) named "speed" that represents a
232 counter. This counter is read every five minutes (this is the default
233 therefore you don't have to put "--step=300"). In the same database
234 two round robin archives (RRAs) are kept, one averages the data every
235 time it is read (e.g., there's nothing to average) and keeps 24 samples
236 (24 times 5 minutes is 2 hours). The other averages 6 values (half
237 hour) and contains 10 such averages (e.g. 5 hours).
239 RRDtool works with special time stamps coming from the UNIX world.
240 This time stamp is the number of seconds that passed since January 1st
241 1970 UTC. The time stamp value is translated into local time and it
242 will therefore look different for different time zones.
244 Chances are that you are not in the same part of the world as I am.
245 This means your time zone is different. In all examples where I talk
246 about time, the hours may be wrong for you. This has little effect on
247 the results of the examples, just correct the hours while reading. As
248 an example: where I will see "12:05" the UK folks will see "11:05".
250 We now have to fill our database with some numbers. We'll pretend to
251 have read the following numbers:
253 12:05 12345 km
254 12:10 12357 km
255 12:15 12363 km
256 12:20 12363 km
257 12:25 12363 km
258 12:30 12373 km
259 12:35 12383 km
260 12:40 12393 km
261 12:45 12399 km
262 12:50 12405 km
263 12:55 12411 km
264 13:00 12415 km
265 13:05 12420 km
266 13:10 12422 km
267 13:15 12423 km
269 We fill the database as follows:
271 rrdtool update test.rrd 920804700:12345 920805000:12357 920805300:12363
272 rrdtool update test.rrd 920805600:12363 920805900:12363 920806200:12373
273 rrdtool update test.rrd 920806500:12383 920806800:12393 920807100:12399
274 rrdtool update test.rrd 920807400:12405 920807700:12411 920808000:12415
275 rrdtool update test.rrd 920808300:12420 920808600:12422 920808900:12423
277 This reads: update our test database with the following numbers
279 time 920804700, value 12345
280 time 920805000, value 12357
282 etcetera.
284 As you can see, it is possible to feed more than one value into the
285 database in one command. I had to stop at three for readability but the
286 real maximum per line is OS dependent.
288 We can now retrieve the data from our database using "rrdtool fetch":
290 rrdtool fetch test.rrd AVERAGE --start 920804400 --end 920809200
292 It should return the following output:
294 speed
296 920804700: nan
297 920805000: 4.0000000000e-02
298 920805300: 2.0000000000e-02
299 920805600: 0.0000000000e+00
300 920805900: 0.0000000000e+00
301 920806200: 3.3333333333e-02
302 920806500: 3.3333333333e-02
303 920806800: 3.3333333333e-02
304 920807100: 2.0000000000e-02
305 920807400: 2.0000000000e-02
306 920807700: 2.0000000000e-02
307 920808000: 1.3333333333e-02
308 920808300: 1.6666666667e-02
309 920808600: 6.6666666667e-03
310 920808900: 3.3333333333e-03
311 920809200: nan
313 If it doesn't, something may be wrong. Perhaps your OS will print
314 "NaN" in a different form. "NaN" stands for "Not A Number". If your OS
315 writes "U" or "UNKN" or something similar that's okay. If something
316 else is wrong, it will probably be due to an error you made (assuming
317 that my tutorial is correct of course :-). In that case: delete the
318 database and try again.
320 The meaning of the above output will become clear below.
322 T\bTi\bim\bme\be t\bto\bo c\bcr\bre\bea\bat\bte\be s\bso\bom\bme\be g\bgr\bra\bap\bph\bhi\bic\bcs\bs
323 Try the following command:
325 rrdtool graph speed.png \
326 --start 920804400 --end 920808000 \
327 DEF:myspeed=test.rrd:speed:AVERAGE \
328 LINE2:myspeed#FF0000
330 This will create speed.png which starts at 12:00 and ends at 13:00.
331 There is a definition of a variable called myspeed, using the data from
332 RRA "speed" out of database "test.rrd". The line drawn is 2 pixels high
333 and represents the variable myspeed. The color is red (specified by its
334 rgb-representation, see below).
336 You'll notice that the start of the graph is not at 12:00 but at 12:05.
337 This is because we have insufficient data to tell the average before
338 that time. This will only happen when you miss some samples, this will
339 not happen a lot, hopefully.
341 If this has worked: congratulations! If not, check what went wrong.
343 The colors are built up from red, green and blue. For each of the
344 components, you specify how much to use in hexadecimal where 00 means
345 not included and FF means fully included. The "color" white is a
346 mixture of red, green and blue: FFFFFF The "color" black is all colors
347 off: 000000
349 red #FF0000
350 green #00FF00
351 blue #0000FF
352 magenta #FF00FF (mixed red with blue)
353 gray #555555 (one third of all components)
355 Additionally you can (with a recent RRDtool) add an alpha channel
356 (transparency). The default will be "FF" which means non-transparent.
358 The PNG you just created can be displayed using your favorite image
359 viewer. Web browsers will display the PNG via the URL
360 "file:///the/path/to/speed.png"
363 When looking at the image, you notice that the horizontal axis is
364 labeled 12:10, 12:20, 12:30, 12:40 and 12:50. Sometimes a label doesn't
365 fit (12:00 and 13:00 would be likely candidates) so they are skipped.
367 The vertical axis displays the range we entered. We provided kilometers
368 and when divided by 300 seconds, we get very small numbers. To be
369 exact, the first value was 12 (12357-12345) and divided by 300 this
370 makes 0.04, which is displayed by RRDtool as "40 m" meaning "40/1000".
371 The "m" (milli) has nothing to do with meters (also m), kilometers or
372 millimeters! RRDtool doesn't know about the physical units of our data,
373 it just works with dimensionless numbers.
375 If we had measured our distances in meters, this would have been
376 (12357000-12345000)/300 = 12000/300 = 40.
378 As most people have a better feel for numbers in this range, we'll
379 correct that. We could recreate our database and store the correct
380 data, but there is a better way: we do some calculations while creating
381 the png file!
383 rrdtool graph speed2.png \
384 --start 920804400 --end 920808000 \
385 --vertical-label m/s \
386 DEF:myspeed=test.rrd:speed:AVERAGE \
387 CDEF:realspeed=myspeed,1000,\* \
388 LINE2:realspeed#FF0000
390 Note: I need to escape the multiplication operator * with a backslash.
391 If I don't, the operating system may interpret it and use it for file
392 name expansion. You could also place the line within quotation marks
393 like so:
395 "CDEF:realspeed=myspeed,1000,*" \
397 It boils down to: it is RRDtool which should see *, not your shell.
398 And it is your shell interpreting \, not RRDtool. You may need to
399 adjust examples accordingly if you happen to use an operating system or
400 shell which behaves differently.
402 After viewing this PNG, you notice the "m" (milli) has disappeared.
403 This it what the correct result would be. Also, a label has been added
404 to the image. Apart from the things mentioned above, the PNG should
405 look the same.
407 The calculations are specified in the CDEF part above and are in
408 Reverse Polish Notation ("RPN"). What we requested RRDtool to do is:
409 "take the data source myspeed and the number 1000; multiply those".
410 Don't bother with RPN yet, it will be explained later on in more
411 detail. Also, you may want to read my tutorial on CDEFs and Steve
412 Rader's tutorial on RPN. But first finish this tutorial.
414 Hang on! If we can multiply values with 1000, it should also be
415 possible to display kilometers per hour from the same data!
417 To change a value that is measured in meters per second:
419 Calculate meters per hour: value * 3600
420 Calculate kilometers per hour: value / 1000
421 Together this makes: value * (3600/1000) or value * 3.6
423 In our example database we made a mistake and we need to compensate for
424 this by multiplying with 1000. Applying that correction:
426 value * 3.6 * 1000 == value * 3600
428 Now let's create this PNG, and add some more magic ...
430 rrdtool graph speed3.png \
431 --start 920804400 --end 920808000 \
432 --vertical-label km/h \
433 DEF:myspeed=test.rrd:speed:AVERAGE \
434 "CDEF:kmh=myspeed,3600,*" \
435 CDEF:fast=kmh,100,GT,kmh,0,IF \
436 CDEF:good=kmh,100,GT,0,kmh,IF \
437 HRULE:100#0000FF:"Maximum allowed" \
438 AREA:good#00FF00:"Good speed" \
439 AREA:fast#FF0000:"Too fast"
441 Note: here we use another means to escape the * operator by enclosing
442 the whole string in double quotes.
444 This graph looks much better. Speed is shown in km/h and there is even
445 an extra line with the maximum allowed speed (on the road I travel on).
446 I also changed the colors used to display speed and changed it from a
447 line into an area.
449 The calculations are more complex now. For speed measurements within
450 the speed limit they are:
452 Check if kmh is greater than 100 ( kmh,100 ) GT
453 If so, return 0, else kmh ((( kmh,100 ) GT ), 0, kmh) IF
455 For values above the speed limit:
457 Check if kmh is greater than 100 ( kmh,100 ) GT
458 If so, return kmh, else return 0 ((( kmh,100) GT ), kmh, 0) IF
461 I like to believe there are virtually no limits to how RRDtool graph
462 can manipulate data. I will not explain how it works, but look at the
463 following PNG:
465 rrdtool graph speed4.png \
466 --start 920804400 --end 920808000 \
467 --vertical-label km/h \
468 DEF:myspeed=test.rrd:speed:AVERAGE \
469 CDEF:nonans=myspeed,UN,0,myspeed,IF \
470 CDEF:kmh=nonans,3600,* \
471 CDEF:fast=kmh,100,GT,100,0,IF \
472 CDEF:over=kmh,100,GT,kmh,100,-,0,IF \
473 CDEF:good=kmh,100,GT,0,kmh,IF \
474 HRULE:100#0000FF:"Maximum allowed" \
475 AREA:good#00FF00:"Good speed" \
476 AREA:fast#550000:"Too fast" \
477 STACK:over#FF0000:"Over speed"
479 Remember the note in the beginning? I had to remove unknown data from
480 this example. The 'nonans' CDEF is new, and the 6th line (which used to
481 be the 5th line) used to read 'CDEF:kmh=myspeed,3600,*'
483 Let's create a quick and dirty HTML page to view the three PNGs:
485 <HTML><HEAD><TITLE>Speed</TITLE></HEAD><BODY>
486 <IMG src="speed2.png" alt="Speed in meters per second">
487 <BR>
488 <IMG src="speed3.png" alt="Speed in kilometers per hour">
489 <BR>
490 <IMG src="speed4.png" alt="Traveled too fast?">
491 </BODY></HTML>
493 Name the file "speed.html" or similar, and look at it in your web
494 browser.
496 Now, all you have to do is measure the values regularly and update the
497 database. When you want to view the data, recreate the PNGs and make
498 sure to refresh them in your browser. (Note: just clicking reload may
499 not be enough, especially when proxies are involved. Try shift-reload
500 or ctrl-F5).
503 We've already used the "update" command: it took one or more parameters
504 in the form of "<time>:<value>". You'll be glad to know that you can
505 specify the current time by filling in a "N" as the time. Or you could
506 use the "time" function in Perl (the shortest example in this
507 tutorial):
509 perl -e 'print time, "\n" '
511 How to run a program on regular intervals is OS specific. But here is
512 an example in pseudo code:
514 - Get the value and put it in variable "$speed"
515 - rrdtool update speed.rrd N:$speed
517 (do not try this with our test database, we'll use it in further
518 examples)
520 This is all. Run the above script every five minutes. When you need to
521 know what the graphs look like, run the examples above. You could put
522 them in a script as well. After running that script, view the page
523 index.html we created above.
526 I can imagine very few people that will be able to get real data from
527 their car every five minutes. All other people will have to settle for
528 some other kind of counter. You could measure the number of pages
529 printed by a printer, for example, the cups of coffee made by the
530 coffee machine, a device that counts the electricity used, whatever.
531 Any incrementing counter can be monitored and graphed using the stuff
532 you learned so far. Later on we will also be able to monitor other
533 types of values like temperature.
535 Many people interested in RRDtool will use the counter that keeps track
536 of octets (bytes) transfered by a network device. So let's do just that
537 next. We will start with a description of how to collect data.
539 Some people will make a remark that there are tools which can do this
540 data collection for you. They are right! However, I feel it is
541 important that you understand they are not necessary. When you have to
542 determine why things went wrong you need to know how they work.
544 One tool used in the example has been talked about very briefly in the
545 beginning of this document, it is called SNMP. It is a way of talking
546 to networked equipment. The tool I use below is called "snmpget" and
547 this is how it works:
549 snmpget device password OID
551 or
553 snmpget -v[version] -c[password] device OID
555 For device you substitute the name, or the IP address, of your device.
556 For password you use the "community read string" as it is called in the
557 SNMP world. For some devices the default of "public" might work,
558 however this can be disabled, altered or protected for privacy and
559 security reasons. Read the documentation that comes with your device
560 or program.
562 Then there is this parameter, called OID, which means "object
563 identifier".
565 When you start to learn about SNMP it looks very confusing. It isn't
566 all that difficult when you look at the Management Information Base
567 ("MIB"). It is an upside-down tree that describes data, with a single
568 node as the root and from there a number of branches. These branches
569 end up in another node, they branch out, etc. All the branches have a
570 name and they form the path that we follow all the way down. The
571 branches that we follow are named: iso, org, dod, internet, mgmt and
572 mib-2. These names can also be written down as numbers and are 1 3 6 1
573 2 1.
575 iso.org.dod.internet.mgmt.mib-2 (1.3.6.1.2.1)
577 There is a lot of confusion about the leading dot that some programs
578 use. There is *no* leading dot in an OID. However, some programs can
579 use the above part of OIDs as a default. To indicate the difference
580 between abbreviated OIDs and full OIDs they need a leading dot when you
581 specify the complete OID. Often those programs will leave out the
582 default portion when returning the data to you. To make things worse,
583 they have several default prefixes ...
585 Ok, lets continue to the start of our OID: we had 1.3.6.1.2.1 From
586 there, we are especially interested in the branch "interfaces" which
587 has number 2 (e.g., 1.3.6.1.2.1.2 or 1.3.6.1.2.1.interfaces).
589 First, we have to get some SNMP program. First look if there is a pre-
590 compiled package available for your OS. This is the preferred way. If
591 not, you will have to get the sources yourself and compile those. The
592 Internet is full of sources, programs etc. Find information using a
593 search engine or whatever you prefer.
595 Assume you got the program. First try to collect some data that is
596 available on most systems. Remember: there is a short name for the part
597 of the tree that interests us most in the world we live in!
599 I will give an example which can be used on Fedora Core 3. If it
600 doesn't work for you, work your way through the manual of snmp and
601 adapt the example to make it work.
603 snmpget -v2c -c public myrouter system.sysDescr.0
605 The device should answer with a description of itself, perhaps an empty
606 one. Until you got a valid answer from a device, perhaps using a
607 different "password", or a different device, there is no point in
608 continuing.
610 snmpget -v2c -c public myrouter interfaces.ifNumber.0
612 Hopefully you get a number as a result, the number of interfaces. If
613 so, you can carry on and try a different program called "snmpwalk".
615 snmpwalk -v2c -c public myrouter interfaces.ifTable.ifEntry.ifDescr
617 If it returns with a list of interfaces, you're almost there. Here's
618 an example:
619 [user@host /home/alex]$ snmpwalk -v2c -c public cisco 2.2.1.2
621 interfaces.ifTable.ifEntry.ifDescr.1 = "BRI0: B-Channel 1"
622 interfaces.ifTable.ifEntry.ifDescr.2 = "BRI0: B-Channel 2"
623 interfaces.ifTable.ifEntry.ifDescr.3 = "BRI0" Hex: 42 52 49 30
624 interfaces.ifTable.ifEntry.ifDescr.4 = "Ethernet0"
625 interfaces.ifTable.ifEntry.ifDescr.5 = "Loopback0"
627 On this cisco equipment, I would like to monitor the "Ethernet0"
628 interface and from the above output I see that it is number four. I
629 try:
631 [user@host /home/alex]$ snmpget -v2c -c public cisco 2.2.1.10.4 2.2.1.16.4
633 interfaces.ifTable.ifEntry.ifInOctets.4 = 2290729126
634 interfaces.ifTable.ifEntry.ifOutOctets.4 = 1256486519
636 So now I have two OIDs to monitor and they are (in full, this time):
638 1.3.6.1.2.1.2.2.1.10
640 and
642 1.3.6.1.2.1.2.2.1.16
644 both with an interface number of 4.
646 Don't get fooled, this wasn't my first try. It took some time for me
647 too to understand what all these numbers mean. It does help a lot when
648 they get translated into descriptive text... At least, when people are
649 talking about MIBs and OIDs you know what it's all about. Do not
650 forget the interface number (0 if it is not interface dependent) and
651 try snmpwalk if you don't get an answer from snmpget.
653 If you understand the above section and get numbers from your device,
654 continue on with this tutorial. If not, then go back and re-read this
655 part.
658 Let the fun begin. First, create a new database. It contains data from
659 two counters, called input and output. The data is put into archives
660 that average it. They take 1, 6, 24 or 288 samples at a time. They
661 also go into archives that keep the maximum numbers. This will be
662 explained later on. The time in-between samples is 300 seconds, a good
663 starting point, which is the same as five minutes.
665 1 sample "averaged" stays 1 period of 5 minutes
666 6 samples averaged become one average on 30 minutes
667 24 samples averaged become one average on 2 hours
668 288 samples averaged become one average on 1 day
670 Lets try to be compatible with MRTG which stores about the following
671 amount of data:
673 600 5-minute samples: 2 days and 2 hours
674 600 30-minute samples: 12.5 days
675 600 2-hour samples: 50 days
676 732 1-day samples: 732 days
678 These ranges are appended, so the total amount of data stored in the
679 database is approximately 797 days. RRDtool stores the data
680 differently, it doesn't start the "weekly" archive where the "daily"
681 archive stopped. For both archives the most recent data will be near
682 "now" and therefore we will need to keep more data than MRTG does!
684 We will need:
686 600 samples of 5 minutes (2 days and 2 hours)
687 700 samples of 30 minutes (2 days and 2 hours, plus 12.5 days)
688 775 samples of 2 hours (above + 50 days)
689 797 samples of 1 day (above + 732 days, rounded up to 797)
691 rrdtool create myrouter.rrd \
692 DS:input:COUNTER:600:U:U \
693 DS:output:COUNTER:600:U:U \
694 RRA:AVERAGE:0.5:1:600 \
695 RRA:AVERAGE:0.5:6:700 \
696 RRA:AVERAGE:0.5:24:775 \
697 RRA:AVERAGE:0.5:288:797 \
698 RRA:MAX:0.5:1:600 \
699 RRA:MAX:0.5:6:700 \
700 RRA:MAX:0.5:24:775 \
701 RRA:MAX:0.5:288:797
703 Next thing to do is to collect data and store it. Here is an example.
704 It is written partially in pseudo code, you will have to find out what
705 to do exactly on your OS to make it work.
707 while not the end of the universe
708 do
709 get result of
710 snmpget router community 2.2.1.10.4
711 into variable $in
712 get result of
713 snmpget router community 2.2.1.16.4
714 into variable $out
716 rrdtool update myrouter.rrd N:$in:$out
718 wait for 5 minutes
719 done
721 Then, after collecting data for a day, try to create an image using:
723 rrdtool graph myrouter-day.png --start -86400 \
724 DEF:inoctets=myrouter.rrd:input:AVERAGE \
725 DEF:outoctets=myrouter.rrd:output:AVERAGE \
726 AREA:inoctets#00FF00:"In traffic" \
727 LINE1:outoctets#0000FF:"Out traffic"
729 This should produce a picture with one day worth of traffic. One day
730 is 24 hours of 60 minutes of 60 seconds: 24*60*60=86400, we start at
731 now minus 86400 seconds. We define (with DEFs) inoctets and outoctets
732 as the average values from the database myrouter.rrd and draw an area
733 for the "in" traffic and a line for the "out" traffic.
735 View the image and keep logging data for a few more days. If you like,
736 you could try the examples from the test database and see if you can
737 get various options and calculations to work.
739 Suggestion: Display in bytes per second and in bits per second. Make
740 the Ethernet graphics go red if they are over four megabits per second.
743 A few paragraphs back I mentioned the possibility of keeping the
744 maximum values instead of the average values. Let's go into this a bit
745 more.
747 Recall all the stuff about the speed of the car. Suppose we drove at
748 144 km/h during 5 minutes and then were stopped by the police for 25
749 minutes. At the end of the lecture we would take our laptop and create
750 and view the image taken from the database. If we look at the second
751 RRA we did create, we would have the average from 6 samples. The
752 samples measured would be 144+0+0+0+0+0=144, divided by 30 minutes,
753 corrected for the error by 1000, translated into km/h, with a result of
754 24 km/h. I would still get a ticket but not for speeding anymore :)
756 Obviously, in this case we shouldn't look at the averages. In some
757 cases they are handy. If you want to know how many km you had traveled,
758 the averaged picture would be the right one to look at. On the other
759 hand, for the speed that we traveled at, the maximum numbers seen is
760 much more interesting. Later we will see more types.
762 It is the same for data. If you want to know the amount, look at the
763 averages. If you want to know the rate, look at the maximum. Over
764 time, they will grow apart more and more. In the last database we have
765 created, there are two archives that keep data per day. The archive
766 that keeps averages will show low numbers, the archive that shows
767 maxima will have higher numbers.
769 For my car this would translate in averages per day of 96/24=4 km/h (as
770 I travel about 94 kilometers on a day) during working days, and maxima
771 of 120 km/h (my top speed that I reach every day).
773 Big difference. Do not look at the second graph to estimate the
774 distances that I travel and do not look at the first graph to estimate
775 my speed. This will work if the samples are close together, as they are
776 in five minutes, but not if you average.
778 On some days, I go for a long ride. If I go across Europe and travel
779 for 12 hours, the first graph will rise to about 60 km/h. The second
780 one will show 180 km/h. This means that I traveled a distance of 60
781 km/h times 24 h = 1440 km. I did this with a higher speed and a maximum
782 around 180 km/h. However, it probably doesn't mean that I traveled for
783 8 hours at a constant speed of 180 km/h!
785 This is a real example: go with the flow through Germany (fast!) and
786 stop a few times for gas and coffee. Drive slowly through Austria and
787 the Netherlands. Be careful in the mountains and villages. If you would
788 look at the graphs created from the five-minute averages you would get
789 a totally different picture. You would see the same values on the
790 average and maximum graphs (provided I measured every 300 seconds).
791 You would be able to see when I stopped, when I was in top gear, when I
792 drove over fast highways etc. The granularity of the data is much
793 higher, so you can see more. However, this takes 12 samples per hour,
794 or 288 values per day, so it would be a lot of data over a longer
795 period of time. Therefore we average it, eventually to one value per
796 day. From this one value, we cannot see much detail, of course.
798 Make sure you understand the last few paragraphs. There is no value in
799 only a line and a few axis, you need to know what they mean and
800 interpret the data in an appropriate way. This is true for all data.
802 The biggest mistake you can make is to use the collected data for
803 something that it is not suitable for. You would be better off if you
804 didn't have the graph at all.
806 L\bLe\bet\bt'\b's\bs r\bre\bev\bvi\bie\bew\bw w\bwh\bha\bat\bt y\byo\bou\bu n\bno\bow\bw s\bsh\bho\bou\bul\bld\bd k\bkn\bno\bow\bw
807 You know how to create a database and can put data in it. You can get
808 the numbers out again by creating an image, do math on the data from
809 the database and view the result instead of the raw data. You know
810 about the difference between averages and maximum, and when to use
811 which (or at least you should have an idea).
813 RRDtool can do more than what we have learned up to now. Before you
814 continue with the rest of this doc, I recommend that you reread from
815 the start and try some modifications on the examples. Make sure you
816 fully understand everything. It will be worth the effort and helps you
817 not only with the rest of this tutorial, but also in your day to day
818 monitoring long after you read this introduction.
821 All right, you feel like continuing. Welcome back and get ready for an
822 increased speed in the examples and explanations.
824 You know that in order to view a counter over time, you have to take
825 two numbers and divide the difference of them between the time lapsed.
826 This makes sense for the examples I gave you but there are other
827 possibilities. For instance, I'm able to retrieve the temperature from
828 my router in three places namely the inlet, the so called hot-spot and
829 the exhaust. These values are not counters. If I take the difference
830 of the two samples and divide that by 300 seconds I would be asking for
831 the temperature change per second. Hopefully this is zero! If not, the
832 computer room is probably on fire :)
834 So, what can we do? We can tell RRDtool to store the values we measure
835 directly as they are (this is not entirely true but close enough). The
836 graphs we make will look much better, they will show a rather constant
837 value. I know when the router is busy (it works -> it uses more
838 electricity -> it generates more heat -> the temperature rises). I know
839 when the doors are left open (the room is air conditioned) -> the warm
840 air from the rest of the building flows into the computer room -> the
841 inlet temperature rises). Etc. The data type we use when creating the
842 database before was counter, we now have a different data type and thus
843 a different name for it. It is called GAUGE. There are more such data
844 types:
846 - COUNTER we already know this one
847 - GAUGE we just learned this one
848 - DERIVE
849 - ABSOLUTE
851 The two additional types are DERIVE and ABSOLUTE. Absolute can be used
852 like counter with one difference: RRDtool assumes the counter is reset
853 when it's read. That is: its delta is known without calculation by
854 RRDtool whereas RRDtool needs to calculate it for the counter type.
855 Example: our first example (12345, 12357, 12363, 12363) would read:
856 unknown, 12, 6, 0. The rest of the calculations stay the same. The
857 other one, derive, is like counter. Unlike counter, it can also
858 decrease so it can have a negative delta. Again, the rest of the
859 calculations stay the same.
861 Let's try them all:
863 rrdtool create all.rrd --start 978300900 \
864 DS:a:COUNTER:600:U:U \
865 DS:b:GAUGE:600:U:U \
866 DS:c:DERIVE:600:U:U \
867 DS:d:ABSOLUTE:600:U:U \
868 RRA:AVERAGE:0.5:1:10
869 rrdtool update all.rrd \
870 978301200:300:1:600:300 \
871 978301500:600:3:1200:600 \
872 978301800:900:5:1800:900 \
873 978302100:1200:3:2400:1200 \
874 978302400:1500:1:2400:1500 \
875 978302700:1800:2:1800:1800 \
876 978303000:2100:4:0:2100 \
877 978303300:2400:6:600:2400 \
878 978303600:2700:4:600:2700 \
879 978303900:3000:2:1200:3000
880 rrdtool graph all1.png -s 978300600 -e 978304200 -h 400 \
881 DEF:linea=all.rrd:a:AVERAGE LINE3:linea#FF0000:"Line A" \
882 DEF:lineb=all.rrd:b:AVERAGE LINE3:lineb#00FF00:"Line B" \
883 DEF:linec=all.rrd:c:AVERAGE LINE3:linec#0000FF:"Line C" \
884 DEF:lined=all.rrd:d:AVERAGE LINE3:lined#000000:"Line D"
886 R\bRR\bRD\bDt\bto\boo\bol\bl u\bun\bnd\bde\ber\br t\bth\bhe\be M\bMi\bic\bcr\bro\bos\bsc\bco\bop\bpe\be
887 · Line A is a COUNTER type, so it should continuously increment and
888 RRDtool must calculate the differences. Also, RRDtool needs to divide
889 the difference by the amount of time lapsed. This should end up as a
890 straight line at 1 (the deltas are 300, the time is 300).
892 · Line B is of type GAUGE. These are "real" values so they should match
893 what we put in: a sort of a wave.
895 · Line C is of type DERIVE. It should be a counter that can decrease.
896 It does so between 2400 and 0, with 1800 in-between.
898 · Line D is of type ABSOLUTE. This is like counter but it works on
899 values without calculating the difference. The numbers are the same
900 and as you can see (hopefully) this has a different result.
902 This translates in the following values, starting at 23:10 and ending
903 at 00:10 the next day (where "u" means unknown/unplotted):
905 - Line A: u u 1 1 1 1 1 1 1 1 1 u
906 - Line B: u 1 3 5 3 1 2 4 6 4 2 u
907 - Line C: u u 2 2 2 0 -2 -6 2 0 2 u
908 - Line D: u 1 2 3 4 5 6 7 8 9 10 u
910 If your PNG shows all this, you know you have entered the data
911 correctly, the RRDtool executable is working properly, your viewer
912 doesn't fool you, and you successfully entered the year 2000 :)
914 You could try the same example four times, each time with only one of
915 the lines.
917 Let's go over the data again:
919 · Line A: 300,600,900 and so on. The counter delta is a constant 300
920 and so is the time delta. A number divided by itself is always 1
921 (except when dividing by zero which is undefined/illegal).
923 Why is it that the first point is unknown? We do know what we put
924 into the database, right? True, But we didn't have a value to
925 calculate the delta from, so we don't know where we started. It would
926 be wrong to assume we started at zero so we don't!
928 · Line B: There is nothing to calculate. The numbers are as they are.
930 · Line C: Again, the start-out value is unknown. The same story is
931 holds as for line A. In this case the deltas are not constant,
932 therefore the line is not either. If we would put the same numbers in
933 the database as we did for line A, we would have gotten the same
934 line. Unlike type counter, this type can decrease and I hope to show
935 you later on why this makes a difference.
937 · Line D: Here the device calculates the deltas. Therefore we DO know
938 the first delta and it is plotted. We had the same input as with line
939 A, but the meaning of this input is different and thus the line is
940 different. In this case the deltas increase each time with 300. The
941 time delta stays at a constant 300 and therefore the division of the
942 two gives increasing values.
945 There are a few more basics to show. Some important options are still
946 to be covered and we haven't look at counter wraps yet. First the
947 counter wrap: In our car we notice that the counter shows 999987. We
948 travel 20 km and the counter should go to 1000007. Unfortunately, there
949 are only six digits on our counter so it really shows 000007. If we
950 would plot that on a type DERIVE, it would mean that the counter was
951 set back 999980 km. It wasn't, and there has to be some protection for
952 this. This protection is only available for type COUNTER which should
953 be used for this kind of counter anyways. How does it work? Type
954 counter should never decrease and therefore RRDtool must assume it
955 wrapped if it does decrease! If the delta is negative, this can be
956 compensated for by adding the maximum value of the counter + 1. For our
957 car this would be:
959 Delta = 7 - 999987 = -999980 (instead of 1000007-999987=20)
961 Real delta = -999980 + 999999 + 1 = 20
963 At the time of writing this document, RRDtool knows of counters that
964 are either 32 bits or 64 bits of size. These counters can handle the
965 following different values:
967 - 32 bits: 0 .. 4294967295
968 - 64 bits: 0 .. 18446744073709551615
970 If these numbers look strange to you, you can view them in their
971 hexadecimal form:
973 - 32 bits: 0 .. FFFFFFFF
974 - 64 bits: 0 .. FFFFFFFFFFFFFFFF
976 RRDtool handles both counters the same. If an overflow occurs and the
977 delta would be negative, RRDtool first adds the maximum of a small
978 counter + 1 to the delta. If the delta is still negative, it had to be
979 the large counter that wrapped. Add the maximum possible value of the
980 large counter + 1 and subtract the erroneously added small value.
982 There is a risk in this: suppose the large counter wrapped while adding
983 a huge delta, it could happen, theoretically, that adding the smaller
984 value would make the delta positive. In this unlikely case the results
985 would not be correct. The increase should be nearly as high as the
986 maximum counter value for that to happen, so chances are you would have
987 several other problems as well and this particular problem would not
988 even be worth thinking about. Even though, I did include an example, so
989 you can judge for yourself.
991 The next section gives you some numerical examples for counter-wraps.
992 Try to do the calculations yourself or just believe me if your
993 calculator can't handle the numbers :)
995 Correction numbers:
997 - 32 bits: (4294967295 + 1) = 4294967296
998 - 64 bits: (18446744073709551615 + 1)
999 - correction1 = 18446744069414584320
1001 Before: 4294967200
1002 Increase: 100
1003 Should become: 4294967300
1004 But really is: 4
1005 Delta: -4294967196
1006 Correction1: -4294967196 + 4294967296 = 100
1008 Before: 18446744073709551000
1009 Increase: 800
1010 Should become: 18446744073709551800
1011 But really is: 184
1012 Delta: -18446744073709550816
1013 Correction1: -18446744073709550816
1014 + 4294967296 = -18446744069414583520
1015 Correction2: -18446744069414583520
1016 + 18446744069414584320 = 800
1018 Before: 18446744073709551615 ( maximum value )
1019 Increase: 18446744069414584320 ( absurd increase, minimum for
1020 Should become: 36893488143124135935 this example to work )
1021 But really is: 18446744069414584319
1022 Delta: -4294967296
1023 Correction1: -4294967296 + 4294967296 = 0
1024 (not negative -> no correction2)
1026 Before: 18446744073709551615 ( maximum value )
1027 Increase: 18446744069414584319 ( one less increase )
1028 Should become: 36893488143124135934
1029 But really is: 18446744069414584318
1030 Delta: -4294967297
1031 Correction1: -4294967297 + 4294967296 = -1
1032 Correction2: -1 + 18446744069414584320 = 18446744069414584319
1034 As you can see from the last two examples, you need strange numbers for
1035 RRDtool to fail (provided it's bug free of course), so this should not
1036 happen. However, SNMP or whatever method you choose to collect the
1037 data, might also report wrong numbers occasionally. We can't prevent
1038 all errors, but there are some things we can do. The RRDtool "create"
1039 command takes two special parameters for this. They define the minimum
1040 and maximum allowed values. Until now, we used "U", meaning "unknown".
1041 If you provide values for one or both of them and if RRDtool receives
1042 data points that are outside these limits, it will ignore those values.
1043 For a thermometer in degrees Celsius, the absolute minimum is just
1044 under -273. For my router, I can assume this minimum is much higher so
1045 I would set it to 10, where as the maximum temperature I would set to
1046 80. Any higher and the device would be out of order.
1048 For the speed of my car, I would never expect negative numbers and also
1049 I would not expect a speed higher than 230. Anything else, and there
1050 must have been an error. Remember: the opposite is not true, if the
1051 numbers pass this check, it doesn't mean that they are correct. Always
1052 judge the graph with a healthy dose of suspicion if it seems weird to
1053 you.
1056 One important feature of RRDtool has not been explained yet: it is
1057 virtually impossible to collect data and feed it into RRDtool on exact
1058 intervals. RRDtool therefore interpolates the data, so they are stored
1059 on exact intervals. If you do not know what this means or how it works,
1060 then here's the help you seek:
1062 Suppose a counter increases by exactly one for every second. You want
1063 to measure it in 300 seconds intervals. You should retrieve values that
1064 are exactly 300 apart. However, due to various circumstances you are a
1065 few seconds late and the interval is 303. The delta will also be 303 in
1066 that case. Obviously, RRDtool should not put 303 in the database and
1067 make you believe that the counter increased by 303 in 300 seconds.
1068 This is where RRDtool interpolates: it alters the 303 value as if it
1069 would have been stored earlier and it will be 300 in 300 seconds. Next
1070 time you are at exactly the right time. This means that the current
1071 interval is 297 seconds and also the counter increased by 297. Again,
1072 RRDtool interpolates and stores 300 as it should be.
1074 in the RRD in reality
1076 time+000: 0 delta="U" time+000: 0 delta="U"
1077 time+300: 300 delta=300 time+300: 300 delta=300
1078 time+600: 600 delta=300 time+603: 603 delta=303
1079 time+900: 900 delta=300 time+900: 900 delta=297
1081 Let's create two identical databases. I've chosen the time range
1082 920805000 to 920805900 as this goes very well with the example numbers.
1084 rrdtool create seconds1.rrd \
1085 --start 920804700 \
1086 DS:seconds:COUNTER:600:U:U \
1087 RRA:AVERAGE:0.5:1:24
1089 Make a copy
1091 for Unix: cp seconds1.rrd seconds2.rrd
1092 for Dos: copy seconds1.rrd seconds2.rrd
1093 for vms: how would I know :)
1095 Put in some data
1097 rrdtool update seconds1.rrd \
1098 920805000:000 920805300:300 920805600:600 920805900:900
1099 rrdtool update seconds2.rrd \
1100 920805000:000 920805300:300 920805603:603 920805900:900
1102 Create output
1104 rrdtool graph seconds1.png \
1105 --start 920804700 --end 920806200 \
1106 --height 200 \
1107 --upper-limit 1.05 --lower-limit 0.95 --rigid \
1108 DEF:seconds=seconds1.rrd:seconds:AVERAGE \
1109 CDEF:unknown=seconds,UN \
1110 LINE2:seconds#0000FF \
1111 AREA:unknown#FF0000
1112 rrdtool graph seconds2.png \
1113 --start 920804700 --end 920806200 \
1114 --height 200 \
1115 --upper-limit 1.05 --lower-limit 0.95 --rigid \
1116 DEF:seconds=seconds2.rrd:seconds:AVERAGE \
1117 CDEF:unknown=seconds,UN \
1118 LINE2:seconds#0000FF \
1119 AREA:unknown#FF0000
1121 View both images together (add them to your index.html file) and
1122 compare. Both graphs should show the same, despite the input being
1123 different.
1126 It's time now to wrap up this tutorial. We covered all the basics for
1127 you to be able to work with RRDtool and to read the additional
1128 documentation available. There is plenty more to discover about RRDtool
1129 and you will find more and more uses for this package. You can easily
1130 create graphs using just the examples provided and using only RRDtool.
1131 You can also use one of the front ends to RRDtool that are available.
1134 Remember to subscribe to the RRDtool mailing list. Even if you are not
1135 answering to mails that come by, it helps both you and the rest of the
1136 users. A lot of the stuff that I know about MRTG (and therefore about
1137 RRDtool) I've learned while just reading the list without posting to
1138 it. I did not need to ask the basic questions as they are answered in
1139 the FAQ (read it!) and in various mails by other users. With thousands
1140 of users all over the world, there will always be people who ask
1141 questions that you can answer because you read this and other
1142 documentation and they didn't.
1145 The RRDtool manpages
1148 I hope you enjoyed the examples and their descriptions. If you do, help
1149 other people by pointing them to this document when they are asking
1150 basic questions. They will not only get their answers, but at the same
1151 time learn a whole lot more.
1153 Alex van den Bogaerdt <alex@vandenbogaerdt.nl>
1157 1.3.8 2009-04-09 RRDTUTORIAL(1)