Packet loss reported by check_ping?

[Andreas Ericsson]

Jason Byrns wrote:
> Good morning, all.
> 
> I've been tweaking our Nagios install over the years.  I've often 
> wondered exactly how the check_ping plugin reports its results.
> 
> I used to leave this plugin at the default number of packets, which I 
> believe is five.  If and when we saw warnings on packet loss, it would 
> usually give us numbers like 16%, 32%, or multiples like that.  Why not 
> multiples of 20%, given that it tried five packets?
> 

This sounds like a bug in check_ping (in order to accurately report 16% 
packet loss it would have to send at least 6 packets, assuming 16% is 
rounded value). My guess is that it sends one at startup and then one 
each second for five seconds, although I can't verify it because I'm too 
lazy to go look at the code ;-)


> Now I'm worried, though.  I recently changed our default ping commands 
> to only use ONE ping packet.  This was due to advice I found after a 
> Google search, when I was seeing extremely high latencies on my host 
> (and some service) checks.  It certainly seemed to help, too, now my 
> host latency is usually below one second.  (Instead of the FIFTEEN 
> MINUTES it was reporting before that)
> 
> But I still see some reports of packet loss.  Like, 50% is a common 
> number I now see.  HOW is it seeing anything but either 0% or 100% 
> packet loss, if it uses only one packet per host check??
> 

Assuming 0, 50 and 100 are the *only* numbers you see, this tallies with 
my guess above.

> Also, along those lines...  I recently moved from the 1.2 Nagios to the 
> more recent 2.4-2.5 Nagios.  Nagios 1.X seemed to only trigger a host 
> check if a service check failed.  Nagios 2.X seems to ALWAYS run host 
> checks in addition to service checks.  The fact that host checks are 
> currently not run in parallel seems like this can really limit performance.
> 
> Why did that behavior change?

It didn't. Or rather, it wasn't supposed to. I haven't seen anything of 
the kind in any of our ~160 customers installations, so I'm fairly 
certain this doesn't happen for all setups. What does your nagios.log 
report on the ~15 lines prepending the host-check entry?

>  Will host checks be run in parallel for 
> the 3.0 series, like I think I read once...?
> 

Yes, that has been planned.

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

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Nagios-users mailing list
Nagios-users at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nagios-users
::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. 
::: Messages without supporting info will risk being sent to /dev/null