Curtis Doty | 2 Aug 20:14 2003
Picon

smokeping rpm

Howdy all,

I just discovered that a google on smokeping+rpm yields my little aged 
pile of tweaks. So I quickly merged with upstream and rebuilt.

As of this writing, I've one patch
http://www.GreenKey.net/~curtis/rpm/src/SPECPATCH/smokeping-1.23/smokeping-maxhight.patch.html
which seems to partition the apache security better.

And it's merged with the current release. But please note that this 
package takes some ugly shortcuts such as disregarding SpeedyCGI. (In the 
interest of laziness. ;)

It also builds a custom config upon install. The purpose is more academic 
than functional.

Lastly, it makes some style assumptions that can be blamed upon nobody but 
me.

If there are improvements or Tobi's already fixed, I'll gladly try to keep 
things in sync as time permits.

SmokePing 1.23 RPM
http://www.GreenKey.net/~curtis/rpm/noarch/smokeping-1.23-gk1.noarch.rpm
http://www.GreenKey.net/~curtis/rpm/src/smokeping-1.23-gk1.src.rpm

TODO:
 - add SpeedyCGI back in...rpm-style
 - fix broken apache2 conf...still

(Continue reading)

Scott Schappell | 8 Aug 18:56 2003

Sending alerts to alternate addresses on a per target basis?

First, let me apologize if this has been asked before, I did a cursory dig
through the archives and didn't notice anything about this question.

Is it possible to have e-mail sent to different targets on an alert?

Example:

*** Alerts ***
from me <at> mydom.tld
to  me <at> mydom.tld
+ lossdetect
... define alert foo here

*** Targets ***
++ RemoteSite
menu = RemoteSite
title = RemoteSite
alerts = lossdetect
host = 10.0.0.1
alert_email = remote_contact <at> remotesite.tld, me <at> mydom.tld

If this is not possible, I'd like to offer it up as a suggestion for a
future release.  
--

-- 
PGP Key: http://archon.silvertree.org/pgp.txt
"Compassion and retribution are two sides of the same coin. Necessity
dictates on what side the coin will fall."

--
Unsubscribe mailto:smokeping-users-request <at> list.ee.ethz.ch?subject=unsubscribe
(Continue reading)

Dan McGinn-Combs | 8 Aug 20:16 2003

Re: Sending alerts to alternate addresses on a per target basis?

This is of course "doable" a couple of ways.

The way you have recommended would require some changes to the
Smokeping.pm code (not unreasonable) to go along with your recommended
changes to the config file.

The second way to handle this is to setup a script that handles the
decision about where to send messages based on the host that is being
alerted on. There is a point in the Smokeping.pm script which indicates
that if you prefix the "to" address in ***Alerts*** then it will execute
an external script rather than sending an email directly.

Thus:

*** Alerts ***
# Standard Alert
to = |/usr/local/smokeping/bin/tell-someone.pl
from = smokealert <at> mydomain.com

Then in the external script "tell-someone.pl" you can have lines like
this:

sendmail ($from,$fw-alert,$body) if ($ARGV[1] =~ /fw/);
sendmail ($from,$exchange-alert,$body) if ($ARGV[1] =~ /exchange/);
sendmail ($from,$router-alert,$body) if ($ARGV[1] =~ /router/);

where "sendmail" is the Perl code to send an alert and $fw-alert,
$exchange-alert and $router-alert are the e-mail addresses of the folks
managing the respective devices.

(Continue reading)

Edwin D. Vinas | 11 Aug 09:58 2003
Picon

http://10.10.8.34/smokeping/smokeping.cgi We have an error below: ------------------------------------------ Internal Server Error The server encountered an internal error or misconfiguration and was unable to complete your request. Please contact the server administrator, root <at> yanyan.preginet.net and inform them of the time the error occurred, and anything you might have done that may have caused the error. More information about this error may be available in the server error log.


	Apache/1.3.28 Server at yanyan.preginet.net Port 80
	
	----------------------------------------------

 
	When apache error_log is tailed, we have this message:
	------------------------------------------------

 [Mon Aug 11 15:56:06 2003] [error] [client 10.10.8.2] malformed header from
 script. Bad header=Daemonizing /usr/local/smokepi: /usr/local/smokeping/htdocs/smokeping.cgi

 [Mon Aug 11 15:56:06 2003] [error] [client 10.10.8.2] malformed header from
 script. Bad header=Daemonizing /usr/local/smokepi: /usr/local/smokeping/htdocs/smokeping.cgi

 [Mon Aug 11 15:56:07 2003] [error] [client 10.10.8.2] malformed header from
 script. Bad header=Daemonizing /usr/local/smokepi: /usr/local/smokeping/htdocs/smokeping.cgi

 ------------------------------------------------
	
	When we run bin/smokeping,
 here is the message:
	------------------------------------------------

 [root <at> yanyan htdocs]# ../bin/smokeping
	### fping seems to report in 1 miliseconds

 Daemonizing ../bin/smokeping ...
	------------------------------------------------

(Continue reading)

Simon Westlake | 18 Aug 21:01 2003
Picon

Smokeping on a single box for large number of hosts

Hello all,

I've read through the previous emails and found a couple of emails touching on this subject, but nothing
that could answer my question once and for all, so I was wondering if anyone could help with this problem.

I'm currently running MRTG to monitor throughput and Smokeping to monitor latency on about 500 devices.
MRTG is daemonized and integrated with RRDtool. On the current box (P3-550/256mb PC133, 7200 RPM IDE
ATA100) the performance leaves much to be desired - the graphs are often not updated in time.

We're looking to monitor at *least* 900 devices right now, and probably 1800 within a year. I'm eyeing a
couple of boxes to run both MRTG and Smokeping on to monitor all these users. Is this even possible, or is it
better to spread the data collection between boxes? It would make our life much easier to do it all on one box
at this stage.

The machine we're looking at right now is:

2.4ghz P4
1GB 533mhz RAM
15K RPM SCSI drive (possibly two in RAID0)

We'd be running Debian as an OS and MRTG/Smokeping at their default intervals (5 mins.) The box wouldn't be
used for anything else other than MRTG/Smokeping at this stage (possibly SLAMon in future.)

Has anyone had experience with something else of this magnitude, and if so, what hardware did you run? Can
anyone comment on the performance we might see from this type of box? On paper, it looks like it should work
alright, but I don't have the equipment to test Smokeping/MRTG extensively enough beforehand, so if
anyone has any information to share, please do.

Thanks for any help you can provide.

(Continue reading)

Tobias Oetiker | 18 Aug 22:03 2003
Picon
Picon

Re: Smokeping on a single box for large number of hosts

Today Simon Westlake wrote:

> Has anyone had experience with something else of this magnitude,
> and if so, what hardware did you run? Can anyone comment on the
> performance we might see from this type of box? On paper, it
> looks like it should work alright, but I don't have the equipment
> to test Smokeping/MRTG extensively enough beforehand, so if
> anyone has any information to share, please do.

Simon,

it is amazing that you have problems with your current box, it
doues not look all that slow ... are you sure your problem is
cpu/disk bound and not latency ? What is the load of the machine ?
Have you tried the forks option ?
tobi
>
> Thanks for any help you can provide.
>
>
> --
> Unsubscribe mailto:smokeping-users-request <at> list.ee.ethz.ch?subject=unsubscribe
> Help        mailto:smokeping-users-request <at> list.ee.ethz.ch?subject=help
> Archive     http://www.ee.ethz.ch/~slist/smokeping-users
> WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi
>

--

-- 
 ______    __   _
/_  __/_  / /  (_) Oetiker  <at>  ISG.EE, ETZ J97, ETH, CH-8092 Zurich
(Continue reading)

Naman Latif | 19 Aug 05:49 2003

Help with telnetIOSPing probe

Hi,
I am trying to implement telnetIOSPing probe for Two different Targets
at the same tree level. Pings will be sourced from the same host.
However smokeping\probe doesn't allow me to define the "PROBE_CONF"
variable twice, which I need to have the second target defined.
My config is as below

++++++++++++++++++++++++++++++++
+++ PROBE_CONF
iospass = abcd
iosuser = efgh
target = x.x.x.x
source = gk.example.com

+++ Target1
menu = whatever
title = whatever
host = x.x.x.x

+++ PROBE_CONF
iospass = abcd
iosuser = efgh
target = y.y.y.y
source = gk.example.com

+++ Target2
menu = whatever1
title = whatever1
host = y.y.y.y
++++++++++++++++++++++++++++++++++++++++++++++++++++
(Continue reading)

Joe Nave | 19 Aug 07:40 2003

Re: Help with telnetIOSPing probe

Naman,

Here is a working example - Hope this helps:

++ A_to_B_Traffic
menu = A -> B
title = A traffic to B
host = w.x.y.z
+++ PROBE_CONF
ioshost = a.b.c.d
iosuser = abc

++ A_to_C_Traffic
menu = A -> C
title = A traffic to C
host = w.x.y.z
+++ PROBE_CONF
ioshost = a.b.c.d
iosuser = abc

++ A_to_D_Traffic
menu = A -> D
title = A traffic to D
host = w.x.y.z
+++ PROBE_CONF
ioshost = a.b.c.d
iosuser = abc

++ A_to_E_Traffic
menu = A -> E
(Continue reading)

Dayo Adeyeye | 19 Aug 10:37 2003
Picon

Installation problems

Hello,

I've installed the rrdtool, fping, speedycgi and 
smokeping. I have also edited the following files : 
smokeping, smokeping.cgi and config .

When I run ./bin/smokeping, I get the following error:

[root <at> voice smokeping-1.20]# ./bin/smokeping
WARNING: Hostname 'www.railway-server.abc' does currently 
not resolve to an IP address
WARNING: Hostname 'www.tiscali-web.abc' does currently not 
resolve to an IP address
WARNING: Hostname 'www.ucl-abc.acc.uk' does currently not 
resolve to an IP address
WARNING: Hostname 'www.gurkoman.ybc' does currently not 
resolve to an IP addressWARNING: Hostname 'www.iu.ali' 
does currently not resolve to an IP address
WARNING: Hostname 'www.berkly.udi' does currently not 
resolve to an IP address
WARNING: Hostname 'www.ucsdddar.art' does currently not 
resolve to an IP addressERROR: FPing 'binary' does not 
point to an executable at (eval 7) line 1

How can I resolve these problems please ?

Kind regards

Dayo

(Continue reading)

Simon Leinen | 19 Aug 11:12 2003
X-Face
Picon

Re: Installation problems

Dayo,

> When I run ./bin/smokeping, I get the following error:

> [root <at> voice smokeping-1.20]# ./bin/smokeping
> WARNING: Hostname 'www.railway-server.abc' does currently 
> not resolve to an IP address
> WARNING: Hostname 'www.tiscali-web.abc' does currently not 
> resolve to an IP address
> WARNING: Hostname 'www.ucl-abc.acc.uk' does currently not 
> resolve to an IP address
[...]

the hostnames in the example configuration are fictitious.  You must
replace them with the names of the actual hosts that you want to
measure the RTTs to.

Personally I would recommend to use hosts that are either of specific
interest to you, or that are lightly used, but running reliably, and
on networks that are of interest to you.  Also, make sure that the
hosts actually respond to ping and that their administrators don't
mind your pinging (the traffic is relatively low, something like 20
PING requests every five minutes - someone please correct me if I'm
wrong).
--

-- 
Simon Leinen				       simon <at> babar.switch.ch
SWITCH				   http://www.switch.ch/misc/leinen/

	       Computers hate being anthropomorphized.

(Continue reading)


Gmane