X-Git-Url: https://git.octo.it/?p=rrdtool.git;a=blobdiff_plain;f=doc%2Frrdgraph.src;h=7f912bed5c5cb7f36eea08aac3cabdbf55fd2837;hp=89b4268453a2ac074b570efa44d8d1022470ad6d;hb=f207955a7e325708d056d3dd912863dc9930a71c;hpb=e7de71c824fd824d345c8e7a99a815fe1525eeef diff --git a/doc/rrdgraph.src b/doc/rrdgraph.src index 89b4268..7f912be 100644 --- a/doc/rrdgraph.src +++ b/doc/rrdgraph.src @@ -1,6 +1,6 @@ =include name -=head1 SYNOPSYS +=head1 SYNOPSIS B I [EB>E ...] @@ -12,7 +12,7 @@ EB>E[ ...] =head1 DESCRIPTION -The B function of B is used to represent the +The B function of B is used to present the data from an B to a human viewer. Its main purpose is to create a nice graphical representation but it can also generate a numerical report. @@ -26,16 +26,15 @@ collect data from two or more databases (one per statement though). If you want to display averages, maxima, percentiles etcetera it is best to collect them now using the -B> statement. At this -stage, this command works at the unprocessed data from the B. -I<(Note: this is not yet true; it works on consolidated information -right now)> +B> statement. +Currently this makes no difference but in a future version +of rrdtool you may want to collect these values before consolidation. The data fetched from the B is then B so that there is exactly one datapoint per pixel in the graph. If you do not take care yourself, B will expand the range slightly -if necessary (in that case the first pixel may very well become -unknown!). +if necessary (in that case the first and/or last pixel may very +well become unknown!). Sometimes data is not exactly as you would like to display it. For instance, you might be collecting B per second but want to @@ -43,24 +42,19 @@ display B per second. This is where the B> command is designed for. After B the data, a copy is made and this copy is modified using a rather flexible B> command -set. If you use B> -statements after this, they work on the consolidated data and may -return other values for maximum, minimum etcetera! +set. When you are done fetching and processing the data, it is time to graph it (or print it). This ends the B sequence. =head1 OPTIONS -It is expected that most options will move to the graph definition -statements (after all, most of them do define graph elements...). - =over 4 =item filename The name and path of the graph to generate. It is recommended to -end this in C<.png>, C<.svg> or C<.eps> but B does not enforce this. +end this in C<.png>, C<.svg> or C<.eps> but B does not enforce this. I can be 'C<->' to send the image to C. In that case, no other output is generated. @@ -71,7 +65,7 @@ Z<> B<[-s|--start EtimeE]> B<[-e|--end EtimeE]> -B<[--step EsecondsE]> +B<[-S|--step EsecondsE]> The start and end of what you would like to display, and which B the data should come from. Defaults are: 1 day ago until @@ -93,25 +87,27 @@ B<[-t|--title EstringE]> B<[-v|--vertical-label EstringE]> A horizontal string at the top of the graph and/or a vertically -placed string at the left hand side of the graph. I The string can contain formatter options that -are used to include variables (from Bs) and newlines. +placed string at the left hand side of the graph. Z<> =item Size B<[-w|--width EpixelsE]> -B<[-h|--heigth EpixelsE]> +B<[-h|--height EpixelsE]> +B<[-j|--only-graph]> The width and height of the B (the part of the graph with the actual lines and such). Defaults are 400 pixels by 100 pixels. +If you specify the B<--only-graph> option and set the height E 32 +pixels you will get a tiny graph image to use as an icon in a potential +overview. All labeling will be stripped off the graph. + Z<> =item Limits -I B<[-u|--upper-limit EvalueE]> B<[-l|--lower-limit EvalueE]> B<[-r|--rigid]> @@ -123,37 +119,7 @@ at least from B to B. Autoscaling will still permit those boundaries to be stretched unless the B option is set. -I -B<[--maximum-upper-limit EvalueE]> -B<[--minimum-upper-limit EvalueE]> -B<[--maximum-lower-limit EvalueE]> -B<[--minimum-lower-limit EvalueE]> - -By default the graph will be autoscaling so that it displays the -portion of the y-axis that is actually used. You can change this -behaviour by setting the limits. The displayed y-axis will show -at most B and at least B -at the top, and similarily at least B and -at most B at the bottom. The default is to -display at most B (so: no limit) and at least -B (no minimal value) at the top. The bottom of -the graph has similar defaults. Note that the minimum lower limit -is the lowest one so you should compare this with maximum upper -limit when you try to figure out what you should set. - -To make sure the graph shows the range of I<-1000> to I<2000>, -optionally expanding to no more than I<-3000> to I<4000>, -set the following options: - ---maximum-upper-limit 4000 --minimum-upper-limit 2000 ---maximum-lower-limit -1000 --minimum-lower-limit -3000 - -To mimic the old B option, you can do: - ---maximum-upper-limit 4000 --minimum-upper-limit 4000 ---maximum-lower-limit -3000 --minimum-lower-limit -3000 - -B<[--alt-autoscale]> +B<[-A|--alt-autoscale]> Sometimes the default algorithm for selecting the y-axis scale is not performing very well. Normally the scale is selected from a predefined @@ -163,7 +129,7 @@ maximum y-axis from the actual minimum and maximum values. Our example would display slightly less than C<260-0.001> to slightly more than C<260+0.001> (Contributed by Sasha Mikheev). -B<[--alt-autoscale-max]> +B<[-M|--alt-autoscale-max]> Where C<--alt-autoscale> will modify both the absolute maximum AND minimum values, this option will only affect the maximum value. The minimum @@ -171,11 +137,11 @@ value, if not defined on the command line, will be 0. This option can be useful when graphing router traffic when the WAN line uses compression, and thus the throughput may be higher than the WAN line speed. -B<[--no-gridfit]> +B<[-N|--no-gridfit]> -To avoid antialiasing effects gridlines are placed on +To avoid anti-aliasing effects gridlines are placed on integer pixel values. This is by default done by extending -the scale so gridlines happends to be spaced using an +the scale so gridlines happens to be spaced using an integer number of pixels, and starts on integer pixel value. This might extend the scale too much for some logarithmic scales and for linear scales where --alt-autoscale is needed. @@ -210,7 +176,7 @@ in I and a I format string in I. I defines where each label will be placed. If it is zero, the label will be placed right under the corresponding line (useful for hours, dates etcetera). If you specify a number of seconds here the label is -centered in this interval (useful for monday, januari etcetera). +centered in this interval (useful for Monday, January etcetera). Example: C<--x-grid MINUTE:10:HOUR:1:HOUR:4:0:%X> @@ -237,7 +203,7 @@ placed every I