[Nagiosplug-help] check_icmp oddness

Andreas Ericsson ae at op5.se
Mon Apr 21 16:42:07 CEST 2008

Israel Brewster wrote:
> On Apr 9, 2008, at 12:18 AM, Andreas Ericsson wrote:
>> Israel Brewster wrote:
>>> On Apr 4, 2008, at 1:42 AM, Andreas Ericsson wrote:
>>>> Israel Brewster wrote:
>>>>> I am having some difficulty using the check_icmp plugin to monitor 
>>>>> a  number of my hosts, specifically Linksys RV082 routers running 
>>>>> the  latest firmware. What's happening is if I have a command line 
>>>>> such as:
>>>>> ./check_icmp -n 2 -m 1 -w 3000,80% -c 5000,100% xx.xxx.xxx.111  
>>>>> xx.xxx.xxx.33
>>>> Add '-i 0.5s' to the command line and it might just magically start
>>>> working. That will increase the packet interval, which can sometimes
>>>> be too short when sending to a single host with multiple IP's.
>>>> Especially if there's a firewall in between.
>>> Well, that partially worked. If I increase that value to a 
>>> ridiculously high amount (>10s) the packet loss drops to only about 
>>> 33% (with 5 pings) rather than the 60% I get with the default 
>>> settings. The pattern, however, is the same- only the first x packets 
>>> received, where x is a constant in relation to the number of packets 
>>> sent. One thing I find a bit confusing- the help text lists the -i 
>>> switch as the "MAX packet interval". What determines the min? I would 
>>> thing the minimum packet interval would be more of an issue than the 
>>> maximum, although perhaps not considering that setting the maximum 
>>> did make a difference. Thanks again!
>> The minimum packet interval is determined as "if we've already seen
>> replies to all packets we sent to this host, it's probably ready to
>> get another one, so send it out instantly."
> I'm wondering if it isn't this exact logic that is causing my problems. 
> Is there any way to change this? -I does not help- see below.

It's hard to know without knowing exactly which version you're using.
There has been about 20 bugs filed (and fixed) in check_icmp since it
first was released, but unfortunately the nagiosplug project version
and the one we're releasing aren't exactly the same. None of our
customers are having this issue, although we always recommend them to
disable all firewall rules for packets coming from the server we put
there (and things like this doesn't happen unless there's a rate-limiting
firewall causing it).

>> -I (capital i) is used to set a target_interval, which is used as
>> the minimum value to wait between two packets going to the same
>> target.
> The capital i switch does in fact exist on the check_icmp plugin I have, 
> however it is described (as mentioned in another reply to this thread) 
> as MAX target interval, not min.

Looking at the help; yes. Looking at the code; no. This brings us back
to making it excessively hard to fix the problem without knowing exactly
which version you're using.

I'll look into this though and see if I can get my version to forcibly
wait between sending packets without modifying the code. If that works,
it *might* work for you too, otherwise I'll just send you the code for
the one I've got.

Unfortunately I just got home from one weeks vacation, so I'm swanked
in work (more than usual, that is). Poke me again if you haven't heard
more about this come wednesday or so.

Cheers, and thanks for being persistent while at the same time concise
and a generally nice person :)

Andreas Ericsson                   andreas.ericsson at op5.se
OP5 AB                             www.op5.se
Tel: +46 8-230225                  Fax: +46 8-230231

More information about the Help mailing list