[smokeping-users] Optimizing Smokeping - how many tasks can I run

Peter Kristolaitis alter3d at alter3d.ca
Fri Jul 9 19:07:43 CEST 2010


Yup, you essentially have it right, though there a few minor issues... 
see my changes and comments, inline, below.

- Peter

On 09/07/2010 12:59 PM, Mersberger, Robert wrote:
> Hope you don;t  mind but I would like to pass my changes pass you to 
> verify my thought and actions are on the same path.
> From what I understand after reading about multiple probe instances I 
> would like to change my config to the following.  If this is correct 
> thinking how can I verify that it actually is working?
There are a few ways to tell if it's working.  The quick-n-dirty way is 
simply to check the process list on the machine... if you see multiple 
FPing processes, you're good.

You can also spawn SmokePing in foreground mode and watch the output.

> Here are my proposed changes
>
> + FPing
>
> binary = /usr/sbin/fping
>
> ++ FPingSNF_G1
>
> offset = 0%
>
> ++ FPingSNF_G2
>
> offset =25%
>
> ++ FPingOther
>
> offset 50%
>
You're missing the = character on the last probe.

Also, it's not strictly necessary to use offsets at all, though it's not 
a bad idea either.   The offset can be used to 'tweak' stuff so that 
network traffic (and thus network conditions) are a little more even 
across the probe period.

If you're going to use offset, you might want to use 0% / 33% / 66% 
instead to space them evenly unless you have reason to do otherwise.

> .
>
> .
>
> .
>
> *** Targets ***
>
> # Change to new FPing group = SNF_G1
>
> # was probe = FPing
>
> probe = FPingSNF_G1
>
> menu = Top
>
> title = Network Latency Grapher
>
> remark = Welcome to the SmokePing website of Golden Living \
>
> Here you will learn all about latency and packate loss on our network.
>
> + SNFs
>
> menu = SNF's
>
> title = Beverly Skilled Nursing Facilities
>
> alerts = bigloss,someloss,startloss,rttdetect
>
> ++ 00001-00099
>
> menu = 00001-00099
>
> title =SNFs 00001-00099
>
> +++ 00014
>
> menu = 00014
>
> title =00014
>
> host = 00014berva
>
> # Note many other targets follow is and If I understand they will all 
> use FPingSNF_G1 for their probe/
>
> ++ 02200-02399
>
> # Change to new FPing group = SNF_G2
>
> menu = 02200-02399
>
> probe = FPingSNF_G2
>
> title =02200-02399
>
> +++ 02200
>
> menu = 02200
>
> title =02200
>
> host = 02200rapsd
>
> # Note many other targets follow is and If I understand they will all 
> use FPingSNF_G2 for their probe/
>
> + AseraCare
>
> # Change to new FPing group = Other
>
> menu = AseraCare
>
> title =AseraCare
>
> alerts = bigloss,someloss,startloss,rttdetect
>
> ++ 06370
>
> menu = 06370 ASA
>
> probe = FPingOther
>
> title = ASA-06370 172.29.8.65
>
> host = ASA-06370
>
> ++ 06371
>
> menu = 06371 ASA
>
> title =06371 Grand Island, NE ASA 172.29.0.129
>
> host = ASA-06371
>
> # Note many other targets follow is and If I understand they will all 
> use FPingOther for their probe/
>
Actually, all of the following targets will use FPingSNF_G1 -- you've 
misplaced the "probe = " directive.   If you want all hosts under the 
AseraCase tree to use the "Other" probe, it should be in the "+ 
AseraCare" block;  in your example here, only "++ 06370" uses the Other 
probe;  everything else in the AseraCare tree will inherit the default 
from the top level.

Keep in mind that SmokePing's configuration is strictly hierarchical;  
settings pass from parents to children, not from children to siblings.

>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.oetiker.ch/pipermail/smokeping-users/attachments/20100709/6ac0a449/attachment-0001.htm 


More information about the smokeping-users mailing list