1 =for changes please consult me first. Thanks, Alex
5 rrdtutorial - Alex van den Bogaerdt's RRDtool tutorial
9 RRDtool is written by Tobias Oetiker E<lt>tobi@oetiker.chE<gt> with
10 contributions from many people all around the world. This document is
11 written by Alex van den Bogaerdt E<lt>alex@vandenbogaerdt.nlE<gt> 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.
17 It also explains the general things about statistics with a focus on
24 Please don't skip ahead in this document! The first part of this
25 document explains the basics and may be boring. But if you don't
26 understand the basics, the examples will not be as meaningful to you.
28 Sometimes things change. This example used to provide numbers like
29 "0.04" in stead of "4.00000e-02". Those are really the same numbers,
30 just written down differently. Don't be alarmed if a future version
31 of rrdtool displays a slightly different form of output. The examples
32 in this document are correct for version 1.2.0 of RRDtool.
34 Also, sometimes bugs do occur. They may also influence the outcome of
35 the examples. Example speed4.png was suffering from this (the handling
36 of unknown data in an if-statement was wrong). Normal data will be
37 just fine (a bug in rrdtool wouldn't last long) but special cases like
38 NaN, INF and so on may last a bit longer. Try another version if you
39 can, or just live with it.
41 I fixed the speed4.png example (and added a note). There may be other
42 examples which suffer from the same or a similar bug. Try to fix it
43 yourself, which is a great excercise. But please do not submit your
44 result as a fix to the source of this document. Discuss it on the
45 user's list, or write to me.
47 =head2 What is RRDtool?
49 RRDtool refers to Round Robin Database tool.
50 Round robin is a technique that works with a fixed amount of data, and a
51 pointer to the current element. Think of a circle with some dots plotted
52 on the edge. These dots are the places where data can be stored. Draw an
53 arrow from the center of the circle to one of the dots; this is the pointer.
54 When the current data is read or written, the pointer moves to the next
55 element. As we are on a circle there is neither a beginning nor an end, you can
56 go on and on and on. After a while, all the available places will be used and
57 the process automatically reuses old locations. This way, the dataset
58 will not grow in size and therefore requires no maintenance.
59 RRDtool works with with Round Robin Databases (RRDs). It stores and retrieves
62 =head2 What data can be put into an RRD?
64 You name it, it will probably fit as long as it is some sort of
65 time-series data. This means you have to be able to measure some value
66 at several points in time and provide this information to RRDtool. If
67 you can do this, RRDtool will be able to store it. The values must be
68 numerical but don't have to be integers, as is the case with MRTG (the
69 next section will give more details on this more specialized application).
71 Many examples below talk about SNMP which is an acronym for Simple Network
72 Management Protocol. "Simple" refers to the protocol. It does not
73 mean it is simple to manage or monitor a network. After working your
74 way through this document, you should know enough to be able to
75 understand what people are talking about. For now, just realize that
76 SNMP can be used to query devices for the values of counters they keep. It
77 is the value from those counters that we want to store in the RRD.
79 =head2 What can I do with this tool?
81 RRDtool originated from MRTG (Multi Router Traffic Grapher). MRTG
82 started as a tiny little script for graphing the use of a university's
83 connection to the Internet. MRTG was later (ab-)used as a tool for
84 graphing other data sources including temperature, speed, voltage,
85 number of printouts and the like.
87 Most likely you will start to use RRDtool to store and process data
88 collected via SNMP. The data will most likely be bytes (or bits)
89 transferred from and to a network or a computer. But it can also be
90 used to display tidal waves, solar radiation, power consumption,
91 number of visitors at an exhibition, noise levels near an airport,
92 temperature on your favorite holiday location, temperature in the
93 fridge and whatever you imagination can come up with.
95 You only need a sensor to measure the data and be able to feed the
96 numbers into RRDtool. RRDtool then lets you create a database, store
97 data in it, retrieve that data and create graphs in PNG format for
98 display on a web browser. Those PNG images are dependent on the data
99 you collected and could be, for instance, an overview of the average
100 network usage, or the peaks that occurred.
102 =head2 What if I still have problems after reading this document?
104 First of all: read it again! You may have missed something.
105 If you are unable to compile the sources and you have a fairly common
106 OS, it will probably not be the fault of RRDtool. There may be pre-compiled
107 versions around on the Internet. If they come from trusted sources, get
110 If on the other hand the program works but does not give you the
111 expected results, it will be a problem with configuring it. Review
112 your configuration and compare it with the examples that follow.
114 There is a mailing list and an archive of it. Read the list for a few
115 weeks and search the archive. It is considered rude to just ask
116 a question without searching the archives: your problem may already have been
117 solved for somebody else! This is true for most, if not all, mailing lists
118 and not only for this particular one. Look in the documentation that
119 came with RRDtool for the location and usage of the list.
121 I suggest you take a moment to subscribe to the mailing list right now
122 by sending an email to E<lt>rrd-users-request@lists.oetiker.chE<gt> with a
123 subject of "subscribe". If you ever want to leave this list, just write
124 an email to the same address but now with a subject of "unsubscribe".
126 =head2 How will you help me?
128 By giving you some detailed descriptions with detailed examples.
129 I assume that following the instructions in the order presented
130 will give you enough knowledge of RRDtool to experiment for yourself.
131 If it doesn't work the first time, don't give up. Reread the stuff that
132 you did understand, you may have missed something.
134 By following the examples you get some hands-on experience and, even
135 more important, some background information of how it works.
137 You will need to know something about hexadecimal numbers. If you don't
138 then start with reading L<bin_dec_hex> before you continue here.
140 =head2 Your first Round Robin Database
142 In my opinion the best way to learn something is to actually do it.
143 Why not start right now? We will create a database, put some values
144 in it and extract this data again. Your output should be the same
145 as the output that is included in this document.
147 We will start with some easy stuff and compare a car with a router,
148 or compare kilometers (miles if you wish) with bits and bytes. It's
149 all the same: some number over some time.
151 Assume we have a device that transfers bytes to and from the Internet.
152 This device keeps a counter that starts at zero when it is turned on,
153 increasing with every byte that is transferred. This counter will probably have
154 a maximum value. If this value is reached and an extra byte is counted,
155 the counter starts over at zero. This is the same as many counters
156 in the world such as the mileage counter in a car.
158 Most discussions about networking talk about bits per second so lets
159 get used to that right away. Assume a byte is eight bits and start to
160 think in bits not bytes. The counter, however, still counts bytes!
161 In the SNMP world most of the counters are 32 bits. That means they are
162 counting from 0 to 4294967295. We will use these values in the examples.
163 The device, when asked, returns the current value of the counter. We
164 know the time that has passes since we last asked so we now know how
165 many bytes have been transferred ***on average*** per second. This is
166 not very hard to calculate. First in words, then in calculations:
172 Take the current counter, subtract the previous value from it.
176 Do the same with the current time and the previous time (in seconds).
180 Divide the outcome of (1) by the outcome of (2), the result is
181 the amount of bytes per second. Multiply by eight to get the
182 number of bits per second (bps).
186 bps = (counter_now - counter_before) / (time_now - time_before) * 8
188 For some people it may help to translate this to an automobile example.
189 Do not try this example, and if you do, don't blame me for the results!
191 People who are not used to think in kilometers per hour can translate
192 most into miles per hour by dividing km by 1.6 (close enough).
193 I will use the following abbreviations:
196 km: kilometer (= 1000 meters).
199 km/h: kilometers per hour
200 m/s: meters per second
202 You are driving a car. At 12:05 you read the counter in the dashboard
203 and it tells you that the car has moved 12345 km until that moment.
204 At 12:10 you look again, it reads 12357 km. This means you have
205 traveled 12 km in five minutes. A scientist would translate that
206 into meters per second and this makes a nice comparison toward the
207 problem of (bytes per five minutes) versus (bits per second).
209 We traveled 12 kilometers which is 12000 meters. We did that in five
210 minutes or 300 seconds. Our speed is 12000m / 300s or 40 m/s.
212 We could also calculate the speed in km/h: 12 times 5 minutes
213 is an hour, so we have to multiply 12 km by 12 to get 144 km/h.
214 For our native English speaking friends: that's 90 mph so don't
215 try this example at home or where I live :)
217 Remember: these numbers are averages only. There is no way to figure out
218 from the numbers, if you drove at a constant speed. There is an example
219 later on in this tutorial that explains this.
221 I hope you understand that there is no difference in calculating m/s or
222 bps; only the way we collect the data is different. Even the k from kilo
223 is the same as in networking terms k also means 1000.
225 We will now create a database where we can keep all these interesting
226 numbers. The method used to start the program may differ slightly from
227 OS to OS, but I assume you can figure it out if it works different on
228 your's. Make sure you do not overwrite any file on your system when
229 executing the following command and type the whole line as one long
230 line (I had to split it for readability)
231 and skip all of the '\' characters.
233 rrdtool create test.rrd \
235 DS:speed:COUNTER:600:U:U \
236 RRA:AVERAGE:0.5:1:24 \
239 (So enter: C<rrdtool create test.rrd --start 920804400 DS ...>)
241 =head2 What has been created?
243 We created the round robin database called test (test.rrd) which starts at
244 noon the day I started writing this document, 7th of March, 1999 (this date
245 translates to 920804400 seconds as explained below). Our database holds
246 one data source (DS) named "speed" that represents a counter. This counter
247 is read every five minutes (this is the default therefore you don't have to
248 put C<--step=300>). In the same database two round robin archives (RRAs)
249 are kept, one averages the data every time it is read (e.g., there's nothing
250 to average) and keeps 24 samples (24 times 5 minutes is 2 hours). The other
251 averages 6 values (half hour) and contains 10 such averages (e.g. 5 hours).
253 RRDtool works with special time stamps coming from the UNIX world.
254 This time stamp is the number of seconds that passed since January
255 1st 1970 UTC. The time stamp value is translated into local time and
256 it will therefore look different for different time zones.
258 Chances are that you are not in the same part of the world as I am.
259 This means your time zone is different. In all examples where I talk
260 about time, the hours may be wrong for you. This has little effect on
261 the results of the examples, just correct the hours while reading.
262 As an example: where I will see "12:05" the UK folks will see "11:05".
264 We now have to fill our database with some numbers. We'll pretend to
265 have read the following numbers:
283 We fill the database as follows:
285 rrdtool update test.rrd 920804700:12345 920805000:12357 920805300:12363
286 rrdtool update test.rrd 920805600:12363 920805900:12363 920806200:12373
287 rrdtool update test.rrd 920806500:12383 920806800:12393 920807100:12399
288 rrdtool update test.rrd 920807400:12405 920807700:12411 920808000:12415
289 rrdtool update test.rrd 920808300:12420 920808600:12422 920808900:12423
291 This reads: update our test database with the following numbers
293 time 920804700, value 12345
294 time 920805000, value 12357
298 As you can see, it is possible to feed more than one value into the
299 database in one command. I had to stop at three for readability but
300 the real maximum per line is OS dependent.
302 We can now retrieve the data from our database using "rrdtool fetch":
304 rrdtool fetch test.rrd AVERAGE --start 920804400 --end 920809200
306 It should return the following output:
311 920805000: 4.0000000000e-02
312 920805300: 2.0000000000e-02
313 920805600: 0.0000000000e+00
314 920805900: 0.0000000000e+00
315 920806200: 3.3333333333e-02
316 920806500: 3.3333333333e-02
317 920806800: 3.3333333333e-02
318 920807100: 2.0000000000e-02
319 920807400: 2.0000000000e-02
320 920807700: 2.0000000000e-02
321 920808000: 1.3333333333e-02
322 920808300: 1.6666666667e-02
323 920808600: 6.6666666667e-03
324 920808900: 3.3333333333e-03
327 If it doesn't, something may be wrong. Perhaps your OS will print
328 "NaN" in a different form. "NaN" stands for "Not A Number". If your OS
329 writes "U" or "UNKN" or something similar that's okay. If something
330 else is wrong, it will probably be due to an error you made (assuming
331 that my tutorial is correct of course :-). In that case: delete the
332 database and try again.
335 The meaning of the above output will become clear below.
337 =head2 Time to create some graphics
339 Try the following command:
341 rrdtool graph speed.png \
342 --start 920804400 --end 920808000 \
343 DEF:myspeed=test.rrd:speed:AVERAGE \
346 This will create speed.png which starts at 12:00 and ends at 13:00.
347 There is a definition of a variable called myspeed, using the data from RRA
348 "speed" out of database "test.rrd". The line drawn is 2 pixels high
349 and represents the variable myspeed. The color is red (specified by
350 its rgb-representation, see below).
352 You'll notice that the start of the graph is not at 12:00 but at 12:05.
353 This is because we have insufficient data to tell the average before
354 that time. This will only happen when you miss some samples, this will
355 not happen a lot, hopefully.
357 If this has worked: congratulations! If not, check what went wrong.
360 The colors are built up from red, green and blue. For each of the
361 components, you specify how much to use in hexadecimal where 00 means
362 not included and FF means fully included.
363 The "color" white is a mixture of red, green and blue: FFFFFF
364 The "color" black is all colors off: 000000
369 magenta #FF00FF (mixed red with blue)
370 gray #555555 (one third of all components)
372 Additionally you can (with a recent RRDtool) add an alpha channel
373 (transparency). The default will be "FF" which means non-transparent.
375 The PNG you just created can be displayed using your favorite image
376 viewer. Web browsers will display the PNG via the URL
377 "file:///the/path/to/speed.png"
379 =head2 Graphics with some math
381 When looking at the image, you notice that the horizontal axis is labeled
382 12:10, 12:20, 12:30, 12:40 and 12:50. Sometimes a label doesn't fit (12:00
383 and 13:00 would be likely candidates) so they are skipped.
385 The vertical axis displays the range we entered. We provided
386 kilometers and when divided by 300 seconds, we get very small
387 numbers. To be exact, the first value was 12 (12357-12345) and divided
388 by 300 this makes 0.04, which is displayed by RRDtool as "40 m"
389 meaning "40/1000". The "m" (milli) has nothing to do with meters (also m),
390 kilometers or millimeters! RRDtool doesn't know about the physical
391 units of our data, it just works with dimensionless numbers.
393 If we had measured our distances in meters, this would have been
394 (12357000-12345000)/300 = 12000/300 = 40.
396 As most people have a better feel for numbers in this range, we'll
397 correct that. We could recreate our database and store the correct
398 data, but there is a better way: we do some calculations while creating
401 rrdtool graph speed2.png \
402 --start 920804400 --end 920808000 \
403 --vertical-label m/s \
404 DEF:myspeed=test.rrd:speed:AVERAGE \
405 CDEF:realspeed=myspeed,1000,\* \
406 LINE2:realspeed#FF0000
408 Note: I need to escape the multiplication operator * with a backslash.
409 If I don't, the operating system may interpret it and use it for file
410 name expansion. You could also place the line within quotation marks
413 "CDEF:realspeed=myspeed,1000,*" \
415 It boils down to: it is RRDtool which should see *, not your shell.
416 And it is your shell interpreting \, not RRDtool. You may need to
417 adjust examples accordingly if you happen to use an operating
418 system or shell which behaves differently.
420 After viewing this PNG, you notice the "m" (milli) has
421 disappeared. This it what the correct result would be. Also, a label
422 has been added to the image. Apart from the things mentioned above,
423 the PNG should look the same.
425 The calculations are specified in the CDEF part above and are in
426 Reverse Polish Notation ("RPN"). What we requested RRDtool to do is:
427 "take the data source myspeed and the number 1000; multiply
428 those". Don't bother with RPN yet, it will be explained later on in
429 more detail. Also, you may want to read my tutorial on CDEFs and Steve
430 Rader's tutorial on RPN. But first finish this tutorial.
432 Hang on! If we can multiply values with 1000, it should also be possible
433 to display kilometers per hour from the same data!
435 To change a value that is measured in meters per second:
437 Calculate meters per hour: value * 3600
438 Calculate kilometers per hour: value / 1000
439 Together this makes: value * (3600/1000) or value * 3.6
441 In our example database we made a mistake and we need to compensate for
442 this by multiplying with 1000. Applying that correction:
444 value * 3.6 * 1000 == value * 3600
446 Now let's create this PNG, and add some more magic ...
448 rrdtool graph speed3.png \
449 --start 920804400 --end 920808000 \
450 --vertical-label km/h \
451 DEF:myspeed=test.rrd:speed:AVERAGE \
452 "CDEF:kmh=myspeed,3600,*" \
453 CDEF:fast=kmh,100,GT,kmh,0,IF \
454 CDEF:good=kmh,100,GT,0,kmh,IF \
455 HRULE:100#0000FF:"Maximum allowed" \
456 AREA:good#00FF00:"Good speed" \
457 AREA:fast#FF0000:"Too fast"
459 Note: here we use another means to escape the * operator by enclosing
460 the whole string in double quotes.
462 This graph looks much better. Speed is shown in km/h and there is even
463 an extra line with the maximum allowed speed (on the road I travel
464 on). I also changed the colors used to display speed and changed it
465 from a line into an area.
467 The calculations are more complex now. For speed measurements within
468 the speed limit they are:
470 Check if kmh is greater than 100 ( kmh,100 ) GT
471 If so, return 0, else kmh ((( kmh,100 ) GT ), 0, kmh) IF
473 For values above the speed limit:
475 Check if kmh is greater than 100 ( kmh,100 ) GT
476 If so, return kmh, else return 0 ((( kmh,100) GT ), kmh, 0) IF
478 =head2 Graphics Magic
480 I like to believe there are virtually no limits to how RRDtool graph
481 can manipulate data. I will not explain how it works, but look at the
484 rrdtool graph speed4.png \
485 --start 920804400 --end 920808000 \
486 --vertical-label km/h \
487 DEF:myspeed=test.rrd:speed:AVERAGE \
488 CDEF:nonans=myspeed,UN,0,myspeed,IF \
489 CDEF:kmh=nonans,3600,* \
490 CDEF:fast=kmh,100,GT,100,0,IF \
491 CDEF:over=kmh,100,GT,kmh,100,-,0,IF \
492 CDEF:good=kmh,100,GT,0,kmh,IF \
493 HRULE:100#0000FF:"Maximum allowed" \
494 AREA:good#00FF00:"Good speed" \
495 AREA:fast#550000:"Too fast" \
496 STACK:over#FF0000:"Over speed"
498 Remember the note in the beginning? I had to remove unknown data from
499 this example. The 'nonans' CDEF is new, and the 6th line (which used to
500 be the 5th line) used to read 'CDEF:kmh=myspeed,3600,*'
502 Let's create a quick and dirty HTML page to view the three PNGs:
504 <HTML><HEAD><TITLE>Speed</TITLE></HEAD><BODY>
505 <IMG src="speed2.png" alt="Speed in meters per second">
507 <IMG src="speed3.png" alt="Speed in kilometers per hour">
509 <IMG src="speed4.png" alt="Traveled too fast?">
512 Name the file "speed.html" or similar, and look at it in your web browser.
514 Now, all you have to do is measure the values regularly and update the
515 database. When you want to view the data, recreate the PNGs and make
516 sure to refresh them in your browser. (Note: just clicking reload may
517 not be enough, especially when proxies are involved. Try shift-reload
520 =head2 Updates in Reality
522 We've already used the C<update> command: it took one or more
523 parameters in the form of "<time>:<value>". You'll be glad to know
524 that you can specify the current time by filling in a "N" as the time.
525 Or you could use the "time" function in Perl (the shortest example in
528 perl -e 'print time, "\n" '
530 How to run a program on regular intervals is OS specific. But here is
531 an example in pseudo code:
533 - Get the value and put it in variable "$speed"
534 - rrdtool update speed.rrd N:$speed
536 (do not try this with our test database, we'll use it in further examples)
538 This is all. Run the above script every five minutes. When you need to know
539 what the graphs look like, run the examples above. You could put them
540 in a script as well. After running that script, view the page
541 index.html we created above.
543 =head2 Some words on SNMP
545 I can imagine very few people that will be able to get real data from
546 their car every five minutes. All other people will have to settle for
547 some other kind of counter. You could measure the number of pages
548 printed by a printer, for example, the cups of coffee made by the
549 coffee machine, a device that counts the electricity used,
550 whatever. Any incrementing counter can be monitored and graphed using
551 the stuff you learned so far. Later on we will also be able to monitor
552 other types of values like temperature.
554 Many people interested in RRDtool will use the counter that keeps track
555 of octets (bytes) transferred by a network device. So let's do just
556 that next. We will start with a description of how to collect data.
558 Some people will make a remark that there are tools which can do this data
559 collection for you. They are right! However, I feel it is important that
560 you understand they are not necessary. When you have to determine why
561 things went wrong you need to know how they work.
563 One tool used in the example has been talked about very briefly in the
564 beginning of this document, it is called SNMP. It is a way of talking
565 to networked equipment. The tool I use below is called "snmpget" and
566 this is how it works:
568 snmpget device password OID
572 snmpget -v[version] -c[password] device OID
574 For device you substitute the name, or the IP address, of your device.
575 For password you use the "community read string" as it is called in the
576 SNMP world. For some devices the default of "public" might work, however
577 this can be disabled, altered or protected for privacy and security
578 reasons. Read the documentation that comes with your device or program.
580 Then there is this parameter, called OID, which means "object identifier".
582 When you start to learn about SNMP it looks very confusing. It isn't
583 all that difficult when you look at the Management Information Base
584 ("MIB"). It is an upside-down tree that describes data, with a single node
585 as the root and from there a number of branches. These branches end
586 up in another node, they branch out, etc. All the branches have a name
587 and they form the path that we follow all the way down. The branches
588 that we follow are named: iso, org, dod, internet, mgmt and mib-2.
589 These names can also be written down as numbers and are 1 3 6 1 2 1.
591 iso.org.dod.internet.mgmt.mib-2 (1.3.6.1.2.1)
593 There is a lot of confusion about the leading dot that some programs
594 use. There is *no* leading dot in an OID. However, some programs
595 can use the above part of OIDs as a default. To indicate the difference
596 between abbreviated OIDs and full OIDs they need a leading dot when
597 you specify the complete OID. Often those programs will leave out
598 the default portion when returning the data to you. To make things
599 worse, they have several default prefixes ...
601 Ok, lets continue to the start of our OID: we had 1.3.6.1.2.1
602 From there, we are especially interested in the branch "interfaces"
603 which has number 2 (e.g., 1.3.6.1.2.1.2 or 1.3.6.1.2.1.interfaces).
605 First, we have to get some SNMP program. First look if there is a
606 pre-compiled package available for your OS. This is the preferred way.
607 If not, you will have to get the sources yourself and compile those.
608 The Internet is full of sources, programs etc. Find information using
609 a search engine or whatever you prefer.
611 Assume you got the program. First try to collect some data that is
612 available on most systems. Remember: there is a short name for the
613 part of the tree that interests us most in the world we live in!
615 I will give an example which can be used on Fedora Core 3. If it
616 doesn't work for you, work your way through the manual of snmp and
617 adapt the example to make it work.
619 snmpget -v2c -c public myrouter system.sysDescr.0
621 The device should answer with a description of itself, perhaps an
622 empty one. Until you got a valid answer from a device, perhaps using a
623 different "password", or a different device, there is no point in
626 snmpget -v2c -c public myrouter interfaces.ifNumber.0
628 Hopefully you get a number as a result, the number of interfaces.
629 If so, you can carry on and try a different program called "snmpwalk".
631 snmpwalk -v2c -c public myrouter interfaces.ifTable.ifEntry.ifDescr
633 If it returns with a list of interfaces, you're almost there.
635 [user@host /home/alex]$ snmpwalk -v2c -c public cisco 2.2.1.2
637 interfaces.ifTable.ifEntry.ifDescr.1 = "BRI0: B-Channel 1"
638 interfaces.ifTable.ifEntry.ifDescr.2 = "BRI0: B-Channel 2"
639 interfaces.ifTable.ifEntry.ifDescr.3 = "BRI0" Hex: 42 52 49 30
640 interfaces.ifTable.ifEntry.ifDescr.4 = "Ethernet0"
641 interfaces.ifTable.ifEntry.ifDescr.5 = "Loopback0"
643 On this cisco equipment, I would like to monitor the "Ethernet0"
644 interface and from the above output I see that it is number four. I try:
646 [user@host /home/alex]$ snmpget -v2c -c public cisco 2.2.1.10.4 2.2.1.16.4
648 interfaces.ifTable.ifEntry.ifInOctets.4 = 2290729126
649 interfaces.ifTable.ifEntry.ifOutOctets.4 = 1256486519
651 So now I have two OIDs to monitor and they are (in full, this time):
659 both with an interface number of 4.
661 Don't get fooled, this wasn't my first try. It took some time for me too
662 to understand what all these numbers mean. It does help a lot when they
663 get translated into descriptive text... At least, when people are talking
664 about MIBs and OIDs you know what it's all about.
665 Do not forget the interface number (0 if it is not interface dependent)
666 and try snmpwalk if you don't get an answer from snmpget.
668 If you understand the above section and get numbers from your device, continue
669 on with this tutorial. If not, then go back and re-read this part.
671 =head2 A Real World Example
673 Let the fun begin. First, create a new database. It contains data from
674 two counters, called input and output. The data is put into archives
675 that average it. They take 1, 6, 24 or 288 samples at a time.
676 They also go into archives that keep the maximum numbers. This will be
677 explained later on. The time in-between samples is 300 seconds, a good
678 starting point, which is the same as five minutes.
680 1 sample "averaged" stays 1 period of 5 minutes
681 6 samples averaged become one average on 30 minutes
682 24 samples averaged become one average on 2 hours
683 288 samples averaged become one average on 1 day
685 Lets try to be compatible with MRTG which stores about the following
688 600 5-minute samples: 2 days and 2 hours
689 600 30-minute samples: 12.5 days
690 600 2-hour samples: 50 days
691 732 1-day samples: 732 days
693 These ranges are appended, so the total amount of data stored in the
694 database is approximately 797 days. RRDtool stores the data
695 differently, it doesn't start the "weekly" archive where the "daily"
696 archive stopped. For both archives the most recent data will be near
697 "now" and therefore we will need to keep more data than MRTG does!
701 600 samples of 5 minutes (2 days and 2 hours)
702 700 samples of 30 minutes (2 days and 2 hours, plus 12.5 days)
703 775 samples of 2 hours (above + 50 days)
704 797 samples of 1 day (above + 732 days, rounded up to 797)
706 rrdtool create myrouter.rrd \
707 DS:input:COUNTER:600:U:U \
708 DS:output:COUNTER:600:U:U \
709 RRA:AVERAGE:0.5:1:600 \
710 RRA:AVERAGE:0.5:6:700 \
711 RRA:AVERAGE:0.5:24:775 \
712 RRA:AVERAGE:0.5:288:797 \
718 Next thing to do is to collect data and store it. Here is an example.
719 It is written partially in pseudo code, you will have to find out what
720 to do exactly on your OS to make it work.
722 while not the end of the universe
725 snmpget router community 2.2.1.10.4
728 snmpget router community 2.2.1.16.4
731 rrdtool update myrouter.rrd N:$in:$out
736 Then, after collecting data for a day, try to create an image using:
738 rrdtool graph myrouter-day.png --start -86400 \
739 DEF:inoctets=myrouter.rrd:input:AVERAGE \
740 DEF:outoctets=myrouter.rrd:output:AVERAGE \
741 AREA:inoctets#00FF00:"In traffic" \
742 LINE1:outoctets#0000FF:"Out traffic"
744 This should produce a picture with one day worth of traffic.
745 One day is 24 hours of 60 minutes of 60 seconds: 24*60*60=86400, we
746 start at now minus 86400 seconds. We define (with DEFs) inoctets and
747 outoctets as the average values from the database myrouter.rrd and draw
748 an area for the "in" traffic and a line for the "out" traffic.
750 View the image and keep logging data for a few more days.
751 If you like, you could try the examples from the test database and
752 see if you can get various options and calculations to work.
754 Suggestion: Display in bytes per second and in bits per second. Make
755 the Ethernet graphics go red if they are over four megabits per
758 =head2 Consolidation Functions
760 A few paragraphs back I mentioned the possibility of keeping
761 the maximum values instead of the average values. Let's go
762 into this a bit more.
764 Recall all the stuff about the speed of the car. Suppose we drove at 144
765 km/h during 5 minutes and then were stopped by the police for 25 minutes.
766 At the end of the lecture we would take our laptop and create and view the
767 image taken from the database. If we look at the second RRA we did
768 create, we would have the average from 6 samples. The samples measured
769 would be 144+0+0+0+0+0=144, divided by 30 minutes, corrected for the
770 error by 1000, translated into km/h, with a result of 24 km/h.
771 I would still get a ticket but not for speeding anymore :)
773 Obviously, in this case we shouldn't look at the averages. In some
774 cases they are handy. If you want to know how many km you had traveled,
775 the averaged picture would be the right one to look at. On the other hand, for
776 the speed that we traveled at, the maximum numbers seen is much more
777 interesting. Later we will see more types.
779 It is the same for data. If you want to know the amount, look at the
780 averages. If you want to know the rate, look at the maximum.
781 Over time, they will grow apart more and more. In the last database
782 we have created, there are two archives that keep data per day. The
783 archive that keeps averages will show low numbers, the archive that
784 shows maxima will have higher numbers.
786 For my car this would translate in averages per day of 96/24=4 km/h
787 (as I travel about 94 kilometers on a day) during working days, and
788 maxima of 120 km/h (my top speed that I reach every day).
790 Big difference. Do not look at the second graph to estimate the
791 distances that I travel and do not look at the first graph to
792 estimate my speed. This will work if the samples are close together,
793 as they are in five minutes, but not if you average.
795 On some days, I go for a long ride. If I go across Europe and travel
796 for 12 hours, the first graph will rise to about 60 km/h. The second
797 one will show 180 km/h. This means that I traveled a distance of 60
798 km/h times 24 h = 1440 km. I did this with a higher speed and a
799 maximum around 180 km/h. However, it probably doesn't mean that I
800 traveled for 8 hours at a constant speed of 180 km/h!
802 This is a real example: go with the flow through Germany (fast!) and stop
803 a few times for gas and coffee. Drive slowly through Austria and the
804 Netherlands. Be careful in the mountains and villages. If you would
805 look at the graphs created from the five-minute averages you would
806 get a totally different picture. You would see the same values on the
807 average and maximum graphs (provided I measured every 300 seconds).
808 You would be able to see when I stopped, when I was in top gear, when
809 I drove over fast highways etc. The granularity of the data is much
810 higher, so you can see more. However, this takes 12 samples per hour,
811 or 288 values per day, so it would be a lot of data over a longer
812 period of time. Therefore we average it, eventually to one value per
813 day. From this one value, we cannot see much detail, of course.
815 Make sure you understand the last few paragraphs. There is no value
816 in only a line and a few axis, you need to know what they mean and
817 interpret the data in an appropriate way. This is true for all data.
819 The biggest mistake you can make is to use the collected data for
820 something that it is not suitable for. You would be better off if
821 you didn't have the graph at all.
824 =head2 Let's review what you now should know
826 You know how to create a database and can put data in it. You can get
827 the numbers out again by creating an image, do math on the data from
828 the database and view the result instead of the raw data. You know
829 about the difference between averages and maximum, and when to use
830 which (or at least you should have an idea).
832 RRDtool can do more than what we have learned up to now. Before you
833 continue with the rest of this doc, I recommend that you reread from
834 the start and try some modifications on the examples. Make sure you
835 fully understand everything. It will be worth the effort and helps
836 you not only with the rest of this tutorial, but also in your day to day
837 monitoring long after you read this introduction.
839 =head2 Data Source Types
841 All right, you feel like continuing. Welcome back and get ready
842 for an increased speed in the examples and explanations.
844 You know that in order to view a counter over time, you have to
845 take two numbers and divide the difference of them between the
846 time lapsed. This makes sense for the examples I gave you but there
847 are other possibilities. For instance, I'm able to retrieve the
848 temperature from my router in three places namely the inlet, the
849 so called hot-spot and the exhaust. These values are not counters.
850 If I take the difference of the two samples and divide that by
851 300 seconds I would be asking for the temperature change per second.
852 Hopefully this is zero! If not, the computer room is probably on fire :)
854 So, what can we do? We can tell RRDtool to store the values we measure
855 directly as they are (this is not entirely true but close enough). The
856 graphs we make will look much better, they will show a rather constant
857 value. I know when the router is busy (it
858 works -> it uses more electricity -> it generates more heat -> the
859 temperature rises). I know when the doors are left open (the room is
860 air conditioned) -> the warm air from the rest of the building flows into the
861 computer room -> the inlet temperature rises). Etc. The data type we
862 use when creating the database before was counter, we now have a
863 different data type and thus a different name for it. It is called
864 GAUGE. There are more such data types:
866 - COUNTER we already know this one
867 - GAUGE we just learned this one
871 The two additional types are DERIVE and ABSOLUTE. Absolute can be used like
872 counter with one difference: RRDtool assumes the counter is reset when
873 it's read. That is: its delta is known without calculation by RRDtool
874 whereas RRDtool needs to calculate it for the counter type.
875 Example: our first example (12345, 12357, 12363, 12363) would read:
876 unknown, 12, 6, 0. The rest of the calculations stay the same.
877 The other one, derive, is like counter. Unlike counter, it can also
878 decrease so it can have a negative delta. Again, the rest of the
879 calculations stay the same.
883 rrdtool create all.rrd --start 978300900 \
884 DS:a:COUNTER:600:U:U \
886 DS:c:DERIVE:600:U:U \
887 DS:d:ABSOLUTE:600:U:U \
889 rrdtool update all.rrd \
890 978301200:300:1:600:300 \
891 978301500:600:3:1200:600 \
892 978301800:900:5:1800:900 \
893 978302100:1200:3:2400:1200 \
894 978302400:1500:1:2400:1500 \
895 978302700:1800:2:1800:1800 \
896 978303000:2100:4:0:2100 \
897 978303300:2400:6:600:2400 \
898 978303600:2700:4:600:2700 \
899 978303900:3000:2:1200:3000
900 rrdtool graph all1.png -s 978300600 -e 978304200 -h 400 \
901 DEF:linea=all.rrd:a:AVERAGE LINE3:linea#FF0000:"Line A" \
902 DEF:lineb=all.rrd:b:AVERAGE LINE3:lineb#00FF00:"Line B" \
903 DEF:linec=all.rrd:c:AVERAGE LINE3:linec#0000FF:"Line C" \
904 DEF:lined=all.rrd:d:AVERAGE LINE3:lined#000000:"Line D"
906 =head2 RRDtool under the Microscope
912 Line A is a COUNTER type, so it should continuously increment and RRDtool
913 must calculate the differences. Also, RRDtool needs to divide the
914 difference by the amount of time lapsed. This should end up as a
915 straight line at 1 (the deltas are 300, the time is 300).
919 Line B is of type GAUGE. These are "real" values so they should match
920 what we put in: a sort of a wave.
924 Line C is of type DERIVE. It should be a counter that can decrease. It does
925 so between 2400 and 0, with 1800 in-between.
929 Line D is of type ABSOLUTE. This is like counter but it works on
930 values without calculating the difference. The numbers are the same
931 and as you can see (hopefully) this has a different result.
935 This translates in the following values, starting at 23:10 and ending
936 at 00:10 the next day (where "u" means unknown/unplotted):
938 - Line A: u u 1 1 1 1 1 1 1 1 1 u
939 - Line B: u 1 3 5 3 1 2 4 6 4 2 u
940 - Line C: u u 2 2 2 0 -2 -6 2 0 2 u
941 - Line D: u 1 2 3 4 5 6 7 8 9 10 u
943 If your PNG shows all this, you know you have entered the data correctly,
944 the RRDtool executable is working properly, your viewer doesn't fool you,
945 and you successfully entered the year 2000 :)
947 You could try the same example four times, each time with only one of
950 Let's go over the data again:
956 Line A: 300,600,900 and so on. The counter delta is a constant 300 and
957 so is the time delta. A number divided by itself is always 1 (except
958 when dividing by zero which is undefined/illegal).
960 Why is it that the first point is unknown? We do know what we put into
961 the database, right? True, But we didn't have a value to calculate the delta
962 from, so we don't know where we started. It would be wrong to assume we
963 started at zero so we don't!
967 Line B: There is nothing to calculate. The numbers are as they are.
971 Line C: Again, the start-out value is unknown. The same story is holds
972 as for line A. In this case the deltas are not constant, therefore the line
973 is not either. If we would put the same numbers in the database as we did for
974 line A, we would have gotten the same line. Unlike type counter,
975 this type can decrease and I hope to show you later on why
976 this makes a difference.
980 Line D: Here the device calculates the deltas. Therefore we DO know the
981 first delta and it is plotted. We had the same input as with line A, but
982 the meaning of this input is different and thus the line is different.
983 In this case the deltas increase each time with 300. The time delta
984 stays at a constant 300 and therefore the division of the two gives
991 There are a few more basics to show. Some important options are still to
992 be covered and we haven't look at counter wraps yet. First the counter wrap:
993 In our car we notice that the counter shows 999987. We travel 20 km and
994 the counter should go to 1000007. Unfortunately, there are only six digits
995 on our counter so it really shows 000007. If we would plot that on a type
996 DERIVE, it would mean that the counter was set back 999980 km. It wasn't,
997 and there has to be some protection for this. This protection is only
998 available for type COUNTER which should be used for this kind of counter
999 anyways. How does it work? Type counter should never decrease and
1000 therefore RRDtool must assume it wrapped if it does decrease!
1001 If the delta is negative, this can be compensated for by adding the
1002 maximum value of the counter + 1. For our car this would be:
1004 Delta = 7 - 999987 = -999980 (instead of 1000007-999987=20)
1006 Real delta = -999980 + 999999 + 1 = 20
1008 At the time of writing this document, RRDtool knows of counters that
1009 are either 32 bits or 64 bits of size. These counters can handle the
1010 following different values:
1012 - 32 bits: 0 .. 4294967295
1013 - 64 bits: 0 .. 18446744073709551615
1015 If these numbers look strange to you, you can view them in
1016 their hexadecimal form:
1018 - 32 bits: 0 .. FFFFFFFF
1019 - 64 bits: 0 .. FFFFFFFFFFFFFFFF
1021 RRDtool handles both counters the same. If an overflow occurs and
1022 the delta would be negative, RRDtool first adds the maximum of a small
1023 counter + 1 to the delta. If the delta is still negative, it had to be
1024 the large counter that wrapped. Add the maximum possible value of the
1025 large counter + 1 and subtract the erroneously added small value.
1027 There is a risk in this: suppose the large counter wrapped while adding
1028 a huge delta, it could happen, theoretically, that adding the smaller value
1029 would make the delta positive. In this unlikely case the results would
1030 not be correct. The increase should be nearly as high as the maximum
1031 counter value for that to happen, so chances are you would have several
1032 other problems as well and this particular problem would not even be
1033 worth thinking about. Even though, I did include an example, so you
1034 can judge for yourself.
1036 The next section gives you some numerical examples for counter-wraps.
1037 Try to do the calculations yourself or just believe me if your calculator
1038 can't handle the numbers :)
1042 - 32 bits: (4294967295 + 1) = 4294967296
1043 - 64 bits: (18446744073709551615 + 1)
1044 - correction1 = 18446744069414584320
1048 Should become: 4294967300
1051 Correction1: -4294967196 + 4294967296 = 100
1053 Before: 18446744073709551000
1055 Should become: 18446744073709551800
1057 Delta: -18446744073709550816
1058 Correction1: -18446744073709550816
1059 + 4294967296 = -18446744069414583520
1060 Correction2: -18446744069414583520
1061 + 18446744069414584320 = 800
1063 Before: 18446744073709551615 ( maximum value )
1064 Increase: 18446744069414584320 ( absurd increase, minimum for
1065 Should become: 36893488143124135935 this example to work )
1066 But really is: 18446744069414584319
1068 Correction1: -4294967296 + 4294967296 = 0
1069 (not negative -> no correction2)
1071 Before: 18446744073709551615 ( maximum value )
1072 Increase: 18446744069414584319 ( one less increase )
1073 Should become: 36893488143124135934
1074 But really is: 18446744069414584318
1076 Correction1: -4294967297 + 4294967296 = -1
1077 Correction2: -1 + 18446744069414584320 = 18446744069414584319
1079 As you can see from the last two examples, you need strange numbers
1080 for RRDtool to fail (provided it's bug free of course), so this should
1081 not happen. However, SNMP or whatever method you choose to collect the
1082 data, might also report wrong numbers occasionally. We can't prevent all
1083 errors, but there are some things we can do. The RRDtool "create" command
1084 takes two special parameters for this. They define
1085 the minimum and maximum allowed values. Until now, we used "U", meaning
1086 "unknown". If you provide values for one or both of them and if RRDtool
1087 receives data points that are outside these limits, it will ignore those
1088 values. For a thermometer in degrees Celsius, the absolute minimum is
1089 just under -273. For my router, I can assume this minimum is much higher
1090 so I would set it to 10, where as the maximum temperature I would
1091 set to 80. Any higher and the device would be out of order.
1093 For the speed of my car, I would never expect negative numbers and
1094 also I would not expect a speed higher than 230. Anything else,
1095 and there must have been an error. Remember: the opposite is not true,
1096 if the numbers pass this check, it doesn't mean that they are
1097 correct. Always judge the graph with a healthy dose of suspicion if it
1100 =head2 Data Resampling
1102 One important feature of RRDtool has not been explained yet: it is
1103 virtually impossible to collect data and feed it into RRDtool on exact
1104 intervals. RRDtool therefore interpolates the data, so they are stored
1105 on exact intervals. If you do not know what this means or how it
1106 works, then here's the help you seek:
1108 Suppose a counter increases by exactly one for every second. You want
1109 to measure it in 300 seconds intervals. You should retrieve values
1110 that are exactly 300 apart. However, due to various circumstances you
1111 are a few seconds late and the interval is 303. The delta will also be
1112 303 in that case. Obviously, RRDtool should not put 303 in the database
1113 and make you believe that the counter increased by 303 in 300 seconds.
1114 This is where RRDtool interpolates: it alters the 303 value as if it
1115 would have been stored earlier and it will be 300 in 300 seconds.
1116 Next time you are at exactly the right time. This means that the current
1117 interval is 297 seconds and also the counter increased by 297. Again,
1118 RRDtool interpolates and stores 300 as it should be.
1120 in the RRD in reality
1122 time+000: 0 delta="U" time+000: 0 delta="U"
1123 time+300: 300 delta=300 time+300: 300 delta=300
1124 time+600: 600 delta=300 time+603: 603 delta=303
1125 time+900: 900 delta=300 time+900: 900 delta=297
1127 Let's create two identical databases. I've chosen the time range 920805000
1128 to 920805900 as this goes very well with the example numbers.
1130 rrdtool create seconds1.rrd \
1132 DS:seconds:COUNTER:600:U:U \
1133 RRA:AVERAGE:0.5:1:24
1137 for Unix: cp seconds1.rrd seconds2.rrd
1138 for Dos: copy seconds1.rrd seconds2.rrd
1139 for vms: how would I know :)
1143 rrdtool update seconds1.rrd \
1144 920805000:000 920805300:300 920805600:600 920805900:900
1145 rrdtool update seconds2.rrd \
1146 920805000:000 920805300:300 920805603:603 920805900:900
1150 rrdtool graph seconds1.png \
1151 --start 920804700 --end 920806200 \
1153 --upper-limit 1.05 --lower-limit 0.95 --rigid \
1154 DEF:seconds=seconds1.rrd:seconds:AVERAGE \
1155 CDEF:unknown=seconds,UN \
1156 LINE2:seconds#0000FF \
1158 rrdtool graph seconds2.png \
1159 --start 920804700 --end 920806200 \
1161 --upper-limit 1.05 --lower-limit 0.95 --rigid \
1162 DEF:seconds=seconds2.rrd:seconds:AVERAGE \
1163 CDEF:unknown=seconds,UN \
1164 LINE2:seconds#0000FF \
1167 View both images together (add them to your index.html file)
1168 and compare. Both graphs should show the same, despite the
1169 input being different.
1173 It's time now to wrap up this tutorial. We covered all the basics for
1174 you to be able to work with RRDtool and to read the additional
1175 documentation available. There is plenty more to discover about
1176 RRDtool and you will find more and more uses for this package. You can
1177 easily create graphs using just the examples provided and using only
1178 RRDtool. You can also use one of the front ends to RRDtool that are
1183 Remember to subscribe to the RRDtool mailing list. Even if you are not
1184 answering to mails that come by, it helps both you and the rest of the
1185 users. A lot of the stuff that I know about MRTG (and therefore about
1186 RRDtool) I've learned while just reading the list without posting to
1187 it. I did not need to ask the basic questions as they are answered in
1188 the FAQ (read it!) and in various mails by other users. With
1189 thousands of users all over the world, there will always be people who
1190 ask questions that you can answer because you read this and other
1191 documentation and they didn't.
1195 The RRDtool manpages
1199 I hope you enjoyed the examples and their descriptions. If you do, help
1200 other people by pointing them to this document when they are asking
1201 basic questions. They will not only get their answers, but at the same
1202 time learn a whole lot more.
1204 Alex van den Bogaerdt
1205 E<lt>alex@vandenbogaerdt.nlE<gt>
projects / rrdtool.git / blob