[Orca-users] Missing ethernet interfaces - was Re: Strange graphs on Quad-ethernet card

jontankersleyatsabre at yahoo.com jontankersleyatsabre at yahoo.com
Tue Feb 27 08:27:19 PST 2001


I'd considered this option, using the internal orca feature, but I
identified those problems and left them as separate plots.  It WOULD
be nicer in some ways to combine them, but then, if I have 12 QFE
ports, should they ALL be together..... Might lose the ground cover
for the tall trees.

I don't have the maximum rates per card.  And my script doesn't
include all the card types possible.

A cleaner solution might be some sort of loop for all cards dump out
the kind of data that I'm trying to plot now, or some such.  But then
you run into problems with the new Sun GE cards that have double 32bit
ints for the bytes.  And of course you can't just 'concatenate' the
two ints together.  You have to adjust.  I've got a perl script that
does that from the netstat -k data and shoves it into gnuplot.

--- In orca-users at y..., Blair Zajac <blair at a...> wrote:
> Jon,
>
> I just realized the one problem with this patch is that there's no
> way to currently specify different data_max for the different speed
> interfaces.  We could do something like with the data_max increasing
> for each plot:
>
> plot {
> data     1024 * 8 * ((?:(?:be)|(?:elx)|(?:le)|......)\d+)InKB/s
> data     1024 * 8 * $1OuKB/s
> data_max 100000000
> }
>
> plot {
> data     1024 * 8 * ((?:(?:hme)|(?:le)|......)\d+)InKB/s
> data     1024 * 8 * $1OuKB/s
> data_max 1000000000
> }
>
> plot {
> data     1024 * 8 * ((?:(?:ge)|(?:vge)|......)\d+)InKB/s
> data     1024 * 8 * $1OuKB/s
> data_max 10000000000
> }
>
> We would need a list of all different interfaces and the maximum
> bits per second for each one.  Then we would create one plot for
each
> speed interface.
>
> Do you have a list of all Sun interfaces and their speeds?
>
> Regards,
> Blair
>
> Blair Zajac wrote:
> >
> > Jon,
> >
> > After giving it some thought I think the configuration file can
> > be greatly simplified using existing Orca features.  Can people
> > try the following patch to the orcallator.cfg.in file released
> > with Orca 0.26 and let me know if it works.  I tried it on my
> > data and it generated the same plots, with the only different in
> > the default ordering of the Interface plots.
> >
> > This uses Orca's feature to do regular expression matches on
column
> > names to generate plots.  This should remove any need to edit
> > orcallator.cfg any time a new interface is added.
> >
> > Watch the wrapping here on the patch.  Applying this by hand
should
> > be very easy.
> >
> > Regards,
> > Blair
> >
> > --- /u0/blair/Orca/orca-0.26/orcallator/orcallator.cfg.in      
Thu
> > Mar
> > 9 22:25:01 2000
> > +++ ./orcallator.cfg.in Fri Feb 23 23:29:46 2001
> > @@ -229,10 +229,10 @@
> > }
> >
> > plot {
> > -title                  %g Interface Bits Per Second: be0
> > +title                  %g Interface Bits Per Second: $1
> > source                 orcallator
> > -data                   1024 * 8 * be0InKB/s
> > -data                   1024 * 8 * be0OuKB/s
> > +data                   1024 * 8 * (.*\d+)InKB/s
> > +data                   1024 * 8 * $1OuKB/s
> > line_type              area
> > line_type              line1
> > legend                 Input
> > @@ -240,81 +240,6 @@
> > y_legend               Bits/s
> > data_min               0
> > data_max               100000000
> > -href
> >
http://www.gps.caltech.edu/~blair/orca/docs/orcallator.html#interface_
bits_per_second
> > -}
> > -
> > -plot {
> > -title                  %g Interface Bits Per Second: elxl0
> > -source                 orcallator
> > -data                   1024 * 8 * elxl0InKB/s
> > -data                   1024 * 8 * elxl0OuKB/s
> > -line_type              area
> > -line_type              line1
> > -legend                 Input
> > -legend                 Output
> > -y_legend               Bits/s
> > -data_min               0
> > -data_max               100000000
> > -href
> >
http://www.gps.caltech.edu/~blair/orca/docs/orcallator.html#interface_
bits_per_second
> > -}
> > -
> > -plot {
> > -title                  %g Interface Bits Per Second: hme0
> > -source                 orcallator
> > -data                   1024 * 8 * hme0InKB/s
> > -data                   1024 * 8 * hme0OuKB/s
> > -line_type              area
> > -line_type              line1
> > -legend                 Input
> > -legend                 Output
> > -y_legend               Bits/s
> > -data_min               0
> > -data_max               100000000
> > -href
> >
http://www.gps.caltech.edu/~blair/orca/docs/orcallator.html#interface_
bits_per_second
> > -}
> > -
> > -plot {
> > -title                  %g Interface Bits Per Second: hme1
> > -source                 orcallator
> > -data                   1024 * 8 * hme1InKB/s
> > -data                   1024 * 8 * hme1OuKB/s
> > -line_type              area
> > -line_type              line1
> > -legend                 Input
> > -legend                 Output
> > -y_legend               Bits/s
> > -data_min               0
> > -data_max               100000000
> > -href
> >
http://www.gps.caltech.edu/~blair/orca/docs/orcallator.html#interface_
bits_per_second
> > -}
> > -
> > -plot {
> > -title                  %g Interface Bits Per Second: le0
> > -source                 orcallator
> > -data                   1024 * 8 * le0InKB/s
> > -data                   1024 * 8 * le0OuKB/s
> > -line_type              area
> > -line_type              line1
> > -legend                 Input
> > -legend                 Output
> > -y_legend               Bits/s
> > -data_min               0
> > -data_max               10000000
> > -href
> >
http://www.gps.caltech.edu/~blair/orca/docs/orcallator.html#interface_
bits_per_second
> > -}
> > -
> > -plot {
> > -title                  %g Interface Bits Per Second: le1
> > -source                 orcallator
> > -data                   1024 * 8 * le1InKB/s
> > -data                   1024 * 8 * le1OuKB/s
> > -line_type              area
> > -line_type              line1
> > -legend                 Input
> > -legend                 Output
> > -y_legend               Bits/s
> > -data_min               0
> > -data_max               10000000
> > href
> >
http://www.gps.caltech.edu/~blair/orca/docs/orcallator.html#interface_
bits_per_second
> > }
> >
> > jontankersleyatsabre at y... wrote:
> > >
> > > It isn't pretty, but I've added this to our installation, watch
the
> > > wrapping that exists.....
> > > I'm going to send this to Blair directly along with a disk type
> > setup
> > > to build additional plots for attached disks.  The existance of
> > extra
> > > plots that have no data to plot (no interface by that name)
appears
> > to
> > >
> > > be not a problem.  With some extra smarts, only doing those
> > interfaces
> > >
> > > that actually exists instead of a blind add like this, the cfg
file
> > > would be smaller.
> > >
> > > Eliminate the normal network interface stuff (not any
customized)
> > > Add
> > > ### INTERFACES ###
> > > to the end of the script
> > >
> > > Put this script and file in the same base directory as the cfg
file:
> > > makeint.sh
> > > --------------
> > > !/bin/sh
> > > if [ -f $1 ]; then
> > >   mv ${1} ${1}.bak
> > >   ex ${1}.bak <<EOF >/dev/null 2>&1
> > > 1,/### INTERFACES ###/w ${1}
> > > q
> > > EOF
> > >   for i in le0 le1 le2 le3; do
> > >      cat int.cfg | /bin/sed -e "s/XXINTXX/${i}/g" >>${1}
> > >   done
> > >   for i in nf0 nf1 nf2 nf3; do
> > >      cat int.cfg | /bin/sed -e "s/XXINTXX/${i}/g" >>${1}
> > >   done
> > >   for i in ge0 ge1 ge2 ge3; do
> > >      cat int.cfg | /bin/sed -e "s/XXINTXX/${i}/g" >>${1}
> > >   done
> > >   for i in vge0 vge1 vge2 vge3; do
> > >      cat int.cfg | /bin/sed -e "s/XXINTXX/${i}/g" >>${1}
> > >   done
> > > #  for i in 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
20; do
> > >   for i in qfe0 qfe1 qfe2 qfe3 qfe4 qfe5 qfe6 qfe7 qfe8 qfe9
qfe10;
> > do
> > >      cat int.cfg | /bin/sed -e "s/XXINTXX/${i}/g" >>${1}
> > >   done
> > > #  for i in 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
20; do
> > >   for i in hme0 hme1 hme2 hme3 hme4 hme5 hme6 hme7 hme8 hme9
hme10;
> > do
> > >      cat int.cfg | /bin/sed -e "s/XXINTXX/${i}/g" >>${1}
> > >   done
> > > #
> > > # All interfaces...
> > > #
> > > #  for i in 'le(\\d+)' 'nf(\\d+)' 'ge(\\d+)' 'vge(\\d+)'
'qfe(\\d+)'
> > > 'hme(\\d+)'; do
> > > #     abbr=`echo $i | /bin/sed -e "s/(.*)/-all/"`
> > > #     cat int2.cfg | /bin/sed -e "s/XXINTXX/${i}/g" -e
> > > "s/XXTEXTXX/${abbr}/g" >>${1}
> > > #  done
> > > fi
> > > ----------
> > > int.cfg:
> > > ----------
> > >
> > > plot {
> > > title                  %g Interface Bits Per Second: XXINTXX
> > > source                  orcallator
> > > data                  1024 * 8 * XXINTXXInKB/s
> > > data                  1024 * 8 * XXINTXXOuKB/s
> > > line_type            area
> > > line_type            line1
> > > legend                  Input
> > > legend                  Output
> > > y_legend            Bits/s
> > > data_min            0
> > > data_max            100000000
> > > href
> > >
> >
http://www.gps.caltech.edu/~blair/orca/docs/orcallator.html#interface_
> > > bits_per_second
> > > }
> > >
> > > plot {
> > > title                  %g Interface Packets Per Second: XXINTXX
> > > source                  orcallator
> > > data                  XXINTXXIpkt/s
> > > data                  XXINTXXOpkt/s
> > > line_type            area
> > > line_type            line1
> > > legend                  Input
> > > legend                  Output
> > > y_legend            Packets/s
> > > data_min            0
> > > data_max            100000
> > > flush_regexps            1
> > > href
> > >
> >
http://www.gps.caltech.edu/~blair/orca/docs/orcallator.html#interface_
> > > packets_per_second
> > > }
> > >
> > > plot {
> > > title                  %g Interface Errors Per Second: XXINTXX
> > > source                  orcallator
> > > data                  XXINTXXIErr/s
> > > data                  XXINTXXOErr/s
> > > line_type            area
> > > line_type            line1
> > > legend                  Input Errors
> > > legend                  Output Errors
> > > y_legend            Errors/s
> > > data_min            0
> > > flush_regexps            1
> > > href
> > >
> >
http://www.gps.caltech.edu/~blair/orca/docs/orcallator.html#interface_
> > > errors_per_second
> > > }
> > >
> > > plot {
> > > title                  %g Interface Nocanput Rate: XXINTXX
> > > source                  orcallator
> > > data                  XXINTXXNoCP/s
> > > line_type            area
> > > legend                  No Can Put
> > > y_legend            Nocanput/s
> > > data_min            0
> > > flush_regexps            1
> > > href
> > >
> >
http://www.gps.caltech.edu/~blair/orca/docs/orcallator.html#interface_
> > > nocanput_rate
> > > }
> > >
> > > plot {
> > > title                  %g Interface Deferred Packet Rate:
XXINTXX
> > > source                  orcallator
> > > data                  XXINTXXDefr/s
> > > line_type            area
> > > legend                  Deferred Packet
> > > y_legend            Defers/s
> > > data_min            0
> > > flush_regexps            1
> > > href
> > >
> >
http://www.gps.caltech.edu/~blair/orca/docs/orcallator.html#interface_
> > > deferred_packet_rate
> > > }
> > >
> > > plot {
> > > title                  %g Interface Collisions: XXINTXX
> > > source                  orcallator
> > > data                  XXINTXXColl%
> > > line_type            area
> > > legend                  % Collisions
> > > y_legend            Percent
> > > data_min            0
> > > data_max            200
> > > flush_regexps            1
> > > href
> > >
> >
http://www.gps.caltech.edu/~blair/orca/docs/orcallator.html#interface_
> > > collisions
> > > }
> >
> >                    Yahoo! Groups Sponsor
> > [Classmates.com]
> > Click here for Classmates.com
> >
> >
Service.



More information about the Orca-users mailing list