[mrtg] Re: ping-probe questions

[awatthey at mmm.com]

PINGs (ICMP) will always be the first packets to be discarded by any router
when it feels like it.  It would be nice to hear from a router developer to
know exactly when they are thrown away.  I have come to the conlusion that
it does not (usually) mean anything as packets of other types will not be
thrown away.  One cannot assume anything from the number of PINGs that do
not get through.  We have some frame relay links which are especially prone
to PING discards.

I have written a PING perl script that does 4 PINGs with progressively
larger timeouts to get around the above problem (I don't want MRTG hanging
up on any particular SNMP request any longer than possible).  I ignore the
first PING as that may be subject to ARP delays and then I graph the
biggest and smallest values.  Based on the results I see, the smallest
value seems to give gives me a good indication of the link latency.

I'd be interested in any other theories used in PING usage elsewhere.

Alan.


                                                                                                                   
                    John Oliver                                                                                    
                    <joliver at john-o      To:     mrtg at list.ee.ethz.ch                                              
                    liver.net>           cc:     (bcc: Alan Watthey/UK-Europe/3M/US)                               
                                         Subject:     [mrtg] ping-probe questions                                  
                    27/06/2002                                                                                     
                    19:44                                                                                          
                                                                                                                   
                                                                                                                   






I'm using the latest version of ping-probe to monitor several T1s.  All
of the graphs always show a certain amount of packet loss, which simply
isn't correct... if it were, the phones would be ringing off the hook
over here! :-)  What's the story with that?

Also, is there a version of ping-probe which can/will just graph
latency?

--
John Oliver                                    http://www.john-oliver.net/
joliver at john-oliver.net                    http://www.mrtg-monitoring.com/

--
Unsubscribe mailto:mrtg-request at list.ee.ethz.ch?subject=unsubscribe
Archive     http://www.ee.ethz.ch/~slist/mrtg
FAQ         http://faq.mrtg.org    Homepage     http://www.mrtg.org
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi





--
Unsubscribe mailto:mrtg-request at list.ee.ethz.ch?subject=unsubscribe
Archive     http://www.ee.ethz.ch/~slist/mrtg
FAQ         http://faq.mrtg.org    Homepage     http://www.mrtg.org
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi