[RRD] rrdgraph_examples

SYNOPSIS

rrdtool graph /home/httpd/html/test.png --img-format PNG

followed by any of the examples below

DESCRIPTION

For your convenience some of the commands are explained here by using detailed examples. They are not always cut-and-paste ready because comments are intermixed with the examples.

EXAMPLES

Data with multiple resolutions

    --end now --start end-120000s --width 400
    DEF:ds0a=/home/rrdtool/data/router1.rrd:ds0:AVERAGE
    DEF:ds0b=/home/rrdtool/data/router1.rrd:ds0:AVERAGE:step=1800
    DEF:ds0c=/home/rrdtool/data/router1.rrd:ds0:AVERAGE:step=7200
    LINE1:ds0a#0000FF:"default resolution\l"
    LINE1:ds0b#00CCFF:"resolution 1800 seconds per interval\l"
    LINE1:ds0c#FF00FF:"resolution 7200 seconds per interval\l"

Nicely formatted legend section

    DEF:ds0=/home/rrdtool/data/router1.rrd:ds0:AVERAGE
    DEF:ds1=/home/rrdtool/data/router1.rrd:ds1:AVERAGE
    VDEF:ds0max=ds0,MAXIMUM
    VDEF:ds0avg=ds0,AVERAGE
    VDEF:ds0min=ds0,MINIMUM
    VDEF:ds0pct=ds0,95,PERCENT
    VDEF:ds1max=ds1,MAXIMUM
    VDEF:ds1avg=ds1,AVERAGE
    VDEF:ds1min=ds1,MINIMUM
    VDEF:ds1pct=ds1,95,PERCENT

Note: consolidation occurs here.

    CDEF:ds0bits=ds0,8,*
    CDEF:ds1bits=ds1,8,*

Note: 10 spaces to move text to the right

    COMMENT:"          "

Note: the column titles have to be as wide as the columns

    COMMENT:"Maximum    "
    COMMENT:"Average    "
    COMMENT:"Minimum    "

    COMMENT:"95th percentile\l"
    AREA:ds0bits#00C000:"Inbound "
    GPRINT:ds0max:"%6.2lf %Sbps"
    GPRINT:ds0avg:"%6.2lf %Sbps"
    GPRINT:ds0min:"%6.2lf %Sbps"
    GPRINT:ds0pct:"%6.2lf %Sbps\l"
    LINE1:ds1bits#0000FF:"Outbound"
    GPRINT:ds1max:"%6.2lf %Sbps"
    GPRINT:ds1avg:"%6.2lf %Sbps"
    GPRINT:ds1min:"%6.2lf %Sbps"
    GPRINT:ds1pct:"%6.2lf %Sbps\l"

Offsetting a line on the y-axis

Depending on your needs you can do this in two ways:

• Offset the data, then graph this

      DEF:mydata=my.rrd:ds:AVERAGE

  Note: this will also influence any other command that uses "data"

      CDEF:data=mydata,100,+
      LINE1:data#FF0000:"Data with offset"

• Graph the original data, with an offset

      DEF:mydata=my.rrd:ds:AVERAGE

  Note: no color in the first line so it is not visible

      LINE1:100

  Note: the second line gets stacked on top of the first one

      LINE1:data#FF0000:"Data with offset":STACK

Drawing dashed lines

Also works for HRULE and VRULE

• default style: - - - - - LINE1:data#FF0000:"dashed line":dashes
• more fancy style with offset: - - --- - --- - LINE1:data#FF0000:"another dashed line":dashes=15,5,5,10:dash-offset=10

Time ranges

    Last four weeks: --start end-4w --end 00:00
    January 2001:    --start 20010101 --end start+31d
    January 2001:    --start 20010101 --end 20010201
    Last hour:       --start end-1h
    Last 24 hours:   <nothing at all>
    Yesterday:       --end 00:00

Viewing the current and previous week together

    --end now --start end-1w
    DEF:thisweek=router.rrd:ds0:AVERAGE
    DEF:lastweek=router.rrd:ds0:AVERAGE:end=now-1w:start=end-1w

Shift the data forward by one week (604800 seconds)

    SHIFT:lastweek:604800
    [ more of the usual VDEF and CDEF stuff if you like ]
    AREA:lastweek#0000FF:Last\ week
    LINE1:thisweek#FF0000:This\ week

Aberrant Behaviour Detection

If the specialized function RRAs exist for aberrant behavior detection, they can be used to generate the graph of a time series with confidence bands and failures.

   rrdtool graph example.png \
          DEF:obs=monitor.rrd:ifOutOctets:AVERAGE \
          DEF:pred=monitor.rrd:ifOutOctets:HWPREDICT \
          DEF:dev=monitor.rrd:ifOutOctets:DEVPREDICT \
          DEF:fail=monitor.rrd:ifOutOctets:FAILURES \
          TICK:fail#ffffa0:1.0:"Failures\: Average bits out" \
          CDEF:scaledobs=obs,8,* \
          CDEF:upper=pred,dev,2,*,+ \
          CDEF:lower=pred,dev,2,*,- \
          CDEF:scaledupper=upper,8,* \
          CDEF:scaledlower=lower,8,* \
          LINE2:scaledobs#0000ff:"Average bits out" \
          LINE1:scaledupper#ff0000:"Upper Confidence Bound: Average bits out" \
          LINE1:scaledlower#ff0000:"Lower Confidence Bound: Average bits out"

This example generates a graph of the data series in blue (LINE2 with the scaledobs virtual data source), confidence bounds in red (scaledupper and scaledlower virtual data sources), and potential failures (i.e. potential aberrant aberrant behavior) marked by vertical yellow lines (the fail data source).

The raw data comes from an AVERAGE RRA, the finest resolution of the observed time series (one consolidated data point per primary data point). The predicted (or smoothed) values are stored in the HWPREDICT RRA. The predicted deviations (think standard deviation) values are stored in the DEVPREDICT RRA. Finally, the FAILURES RRA contains indicators, with 1 denoting a potential failure.

All of the data is rescaled to bits (instead of Octets) by multiplying by 8. The confidence bounds are computed by an offset of 2 deviations both above and below the predicted values (the CDEFs upper and lower). Vertical lines indicated potential failures are graphed via the TICK graph element, which converts non-zero values in an RRA into tick marks. Here an axis-fraction argument of 1.0 means the tick marks span the entire y-axis, and hence become vertical lines on the graph.

The choice of 2 deviations (a scaling factor) matches the default used internally by the FAILURES RRA. If the internal value is changed (see rrdtune), this graphing command should be changed to be consistent.

A note on data reduction:

The rrdtool graph command is designed to plot data at a specified temporal resolution, regardless of the actually resolution of the data in the RRD file. This can present a problem for the specialized consolidation functions which maintain a one-to-one mapping between primary data points and consolidated data points. If a graph insists on viewing the contents of these RRAs on a coarser temporal scale, the graph command tries to do something intelligent, but the confidence bands and failures no longer have the same meaning and may be misleading.