hobbit monitoring summary report 🔗 link

Hi,
    Is there better way of reporting this?

    Purpose:  to report how many active client and 
              client-less monitoring are done by hobbit.

    My method:
    1. for client installed monitoring, count the number of 
       /home/hobbit/data/rrd/*/vmstat.rrd that are recent
       (ie. newer than a date file created earlier, say 5am)

    2. for client-less monitoring, count the number of 
       latest /home/hobbit/data/histlogs/*/conn/ or
       /home/hobbit/data/histlogs/*/http/ (if both exist,
       then count only one) that are recent (see def. above).
       (some clients may not be pingable; multiple web pages
       on the same server should be counted only once).

Thanks,
       


Yahoo! Canada Toolbar: Search from anywhere on the web, and bookmark your favourite sites. Download it now
http://ca.toolbar.yahoo.com.

Hello,

well you could do a 
ALLCLIENTS=`$HOBBITSERVERHOME/bin/bb 127.0.0.1 "hobbitdboard"|awk -F\| '{print $1}'|sort -u|wc -l` 
to get all clients that actually report to your server.
As second step do a INSTCLIENTS=`$HOBBITSERVERHOME/bin/bb 127.0.0.1 "hobbitdboard test=cpu"|awk -F\| '{print $1}'|sort -u|wc -l` 
to get all clients that have client installed monitoring. I assume if a client is installed then it also reports a cpu status.
Finally do CLIENTLESS=$((ALLCLIENTS-INSTCLIENTS)).

However, clients polled via SNMP are also show the cpu status but have no client installed.
But I hope to have helped you a bit.

Cheers
Torsten

 
Hi,
    Is there better way of reporting this?

    Purpose:  to report how many active client and 
              client-less monitoring are done by hobbit.

    My method:
    1. for client installed monitoring, count the number of 
       /home/hobbit/data/rrd/*/vmstat.rrd that are recent
       (ie. newer than a date file created earlier, say 5am)

    2. for client-less monitoring, count the number of 
       latest /home/hobbit/data/histlogs/*/conn/ or
       /home/hobbit/data/histlogs/*/http/ (if both exist,
       then count only one) that are recent (see def. above).
       (some clients may not be pingable; multiple web pages
       on the same server should be counted only once).

Thanks,
       

Make your browsing faster, safer, and easier with the new Internet Explorer® 8. Optimized for Yahoo! Get it Now for Free!

Hi.

For some customers, we use the html_mail.pl script from Xymonton. But we have problems with the background color being wrong quite often. The subject and icons are correct, but sometimes the background color is, for example, clear or blue when it should be red. Is anyone else using this script and experiencing the same problem? Do you know what the cause might be? I'm no perl guru, but I cannot find any problems in the script.

Regards,
Johan Sjöberg

Thank you for the prompt reply.  Is it possible to distinguish between "Active" clients
those that are promptly responding to tests (say, pings), those that are currently
sending data (say, CPU), and those have not been responding and sending
(maybe the users dispose/shut down the servers without informing to update Hobbit).

My definition of "Active" is if it responds or sends data in the last hour.


RE: [hobbit] hobbit monitoring summary report
To: <user-ae9b8668bcde@xymon.invalid> Subject: RE: [hobbit] hobbit monitoring summary report From: Torsten Richter <user-c862b499d9fa@xymon.invalid> Date: Fri, 21 Aug 2009 09:34:37 +0200 --------------------------------------------------------------------------------

Hello,
well you could do a ALLCLIENTS=`$HOBBITSERVERHOME/bin/bb 127.0.0.1 "hobbitdboard"|awk -F\| '{print $1}'|sort -u|wc -l` to get all clients that actually report to your server.
As second step do a INSTCLIENTS=`$HOBBITSERVERHOME/bin/bb 127.0.0.1 "hobbitdboard test=cpu"|awk -F\| '{print $1}'|sort -u|wc -l` to get all clients that have client installed monitoring. I assume if a client is installed then it also reports a cpu status.
Finally do CLIENTLESS=$((ALLCLIENTS-INSTCLIENTS)).
However, clients polled via SNMP are also show the cpu status but have no client installed.
But I hope to have helped you a bit.
Cheers
Torsten

Looking for the perfect gift? Give the gift of Flickr! 
http://www.flickr.com/gift/

My explanation for the differences below is that hobbitdboard
is reporting inactive clients as well as active, while A and B
look at only active rrd files (recent changed date).
 
What do you think?  Is there another rrd, beside vmstat.rrd, that is good for checking
CPU?

PS: I'm using the unix comm command:
C = comm -2 -3 A B | wc -l
F = comm -2 -3 D E | wc -l 

A = all active tcp.conn.rrd  (229)
B = all active vmstat.rrd    (150)
C = A - (those in both A and B) - (those in B only)
------- Hobbit summary: -------
B.  clients reporting data:  150
C.   client-less,ping-only:  83 
D = hobbitdboard test=conn
E = hobbitdboard test=cpu
F = D - (those in both D and E) - (those in E only)
D.    229
E.    164  # should be same as B
F.     69  # should be same as C

--- On Fri, 8/21/09, ye-fee liang <user-1626c7b75791@xymon.invalid> wrote:

From: ye-fee liang <user-1626c7b75791@xymon.invalid>
Subject: Re: hobbit monitoring summary report
To: user-ae9b8668bcde@xymon.invalid
Received: Friday, August 21, 2009, 10:30 AM

Thank you for the prompt reply.  Is it possible to distinguish between "Active" clients
those that are promptly responding to tests (say, pings), those that are currently
sending data (say, CPU), and those have not been responding and sending
(maybe the users dispose/shut down the servers without informing to update Hobbit).
My definition of "Active" is if it responds or sends data in the last hour.

RE: [hobbit] hobbit monitoring summary report
To: <user-ae9b8668bcde@xymon.invalid> 
Subject: RE: [hobbit] hobbit monitoring summary report 

From: Torsten Richter <user-c862b499d9fa@xymon.invalid> 
Date: Fri, 21 Aug 2009 09:34:37 +0200 
Hello,
well you could do a 
ALLCLIENTS=`$HOBBITSERVERHOME/bin/bb 127.0.0.1 "hobbitdboard"|awk -F\| '{print $1}'|sort -u|wc -l` 
to get all clients that actually report to your server.
As second step do a INSTCLIENTS=`$HOBBITSERVERHOME/bin/bb 127.0.0.1 "hobbitdboard test=cpu"|awk -F\| '{print $1}'|sort -u|wc -l` 
to get all clients that have client installed monitoring. I assume if a client is installed then it also reports a cpu status.
Finally do CLIENTLESS=$((ALLCLIENTS-INSTCLIENTS)).
However, clients polled via SNMP are also show the cpu status but have no client installed.
But I hope to have helped you a bit.
Cheers
Torsten 

Ask a question on any topic and get answers from real people. Go to Yahoo! Answers and share what you know at http://ca.answers.yahoo.com

Johan Sjöberg schrieb:

Hi,

change the message type to multipart/related (line 83)

Regards,
Alex

Thanks for the tip.

I have, however, found another problem with the script. The message that is stored in the $BBALPHAMSG is truncated to a limited number of characters. That means, that if the red icon is too far down on the page, the script will not find it, and the color of the mail will be the "worst" one in the message that it receives.

So, is there any way to increase the size limit of the $BBALPHAMSG?

/Johan

-----Original Message-----
From: Alexander Bech [mailto:user-bd67b1e4c445@xymon.invalid] 
Sent: den 25 augusti 2009 11:00
To: user-ae9b8668bcde@xymon.invalid
Subject: Re: [hobbit] Wrong color with html_mail.pl

Johan Sjöberg schrieb:

Hi,

change the message type to multipart/related (line 83)

Regards,
Alex

I'm not sure if this can be done any other way, but you could increase the value of MAX_ALERTMSG_SCRIPTS in hobbitd/do_alert.c then recompile.

Regards,

Malcolm

-- 
Technical copy-editor & proofreader

KDE Proofreading Team
KDE British English Translation Team

http://l10n.kde.org/team-infos.php?teamcode=en_GB

Facebook: http://www.facebook.com/FriendlyAtheist

Jetzt kostenlos herunterladen: Internet Explorer 8 und Mozilla Firefox 3.5 -
sicherer, schneller und einfacher! http://portal.gmx.net/de/go/atbrowser