[hobbit] Custom test results displayed on a test page

[Rafal Roginela]

Hi,

Well thanks to Henrik I got the page to display really cool! I reformatted the display so it could be graphed so now it looks like:

XX.XX.XX.XX:     75
XX.XX.XX.XX:     28
XX.XX.XX.XX:      7
XX.XX.XX.XX:      7
XX.XX.XX.XX:      4

The values are in MB and those are IP's. It is really cool!! Now onto the graph I'm having trouble finding how to use the SPLITNCV that Henrik mentioned. So far I have made the entry for the graph and I am generating the rrd file but it contains data for all hosts in one. I'd like to set the graph definition like the disk utilization where it splits per IP and then it puts it on one graph. 

So far this is as far as I got:
I have the data on the page (thanks Henrik!)
I have inserted "talkers=ncv" into hobbitserver.cfg in the TEST2RRD field
I have inserted a line in hobbitserver.cfg SPLITNCV_talkers="*:GUAGE"

This only generates a single file though talkers.rrd and it looks valid using the rrdtool dump method listed in help. Any help with the graph definition and all that would be very helpful plus I'm sure that I have the SPLITNCV part wrong in defining columns too. 

This is about as far as I can get at this point and there is no graph yet since all I have in hobbitgraph.cfg is:

[talkers]
        TITLE Top Talkers on the VPN
        YAXIS Megabytes

Any help would be appreciated.

Merry Christmas!

Thank You,

Rafal Roginela

Office (847) 827-9740 x109
Fax (847) 493-8031

 This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden.


-----Original Message-----
From: Henrik "Størner [mailto:henrik at hswn.dk] 
Sent: Friday, December 19, 2008 3:49 PM
To: hobbit at hswn.dk
Subject: Re: [hobbit] Custom test results displayed on a test page

In <84AE996EE1457F4DB43A05348E32C33F080E2F at ac-hq-exch.corp.americashloans.net> "Rafal Roginela" <Rafal.Roginela at AmeriCashLoans.net> writes:

>I have a PIX that sends syslog to my hobbit server I also have cron job
>that runs a script that analyses that log file and gives me a nice
>looking results that are piped out to a text file:

>   196 MB XX.XX.XX.XX
>   100 MB XX.XX.XX.XX
>    18 MB XX.XX.XX.XX
>    12 MB XX.XX.XX.XX
>     7 MB XX.XX.XX.XX
>     6 MB XX.XX.XX.XX
>     4 MB XX.XX.XX.XX
>     2 MB XX.XX.XX.XX
>Ran at Fri Dec 19 11:40:36 2008

>What I would like to do is generate a custom test that would be able to,
>at the very least, display this output so I can easily look at it by
>clicking on the green dot for this custom test. 

That's easy. This script will do it, just add a section to hobbitlaunch.cfg
to run this every 5 minutes or so.

  #!/bin/sh

  PIXLOGFILE=/var/log/pixlog.txt
  PIXHOSTNAME=pix.foo.com
  PIXCOLUMNNAME=talkers

  $BB $BBDISP "status $PIXHOSTNAME.$PIXCOLUMNNAME green `date`
  `cat $PIXLOGFILE`
  "

  exit 0

Modify the PIXLOGFILE / PIXHOSTNAME / PIXCOLUMNNAME to suit your
setup.

> But it would be cool if I could graph it 

That's a bit more tricky. I guess those "XX.XX.XX.XX" are IP's ?
And if they are IP-adresses of remote hosts talking to your site,
then perhaps they change a lot ? In that case graphing it is pretty
un-interesting, because one IP will not show up for very long in
your data.

On the other hand, if your PIX only gets traffic from a limited
number of IP's, then it would make sense to create an RRD (graph
database file) for each IP. It would probably be easier if your
output would list the IP first and then the MB-value; then you
could use the SPLITNCV handler that is built into Xymon 4.2.2
to parse the data and generate the RRD files. See the hobbitd_rrd
man-page and "Custom graphs" section in the on-line help on
your Hobbit server.


>PS If anyone has ever ran into this also: My pix is sysloging to the
>hobbit box good; I even have it going to a separate file for easy
>sorting but I'm finding that it clogs up my messages file with duplicate
>entries no mater which facility I try (right now I have used local4
>through local6 changing the appropriate commands in the pix).

See your syslog.conf man-page. It depends on the exact flavor of
syslogd your system has, but on my Linux box one can filter out
certain facility.priority items from a logfile by using the "!"
operator, or by specifying "none" as the priority. So on your
"messages" syslog.conf entry, adding "log4,log5,log6.none" to the 
set of selectors should weed out your PIX logs from going to
the messages file.


Regards,
Henrik


To unsubscribe from the hobbit list, send an e-mail to
hobbit-unsubscribe at hswn.dk