[nsp] Graphing packet loss

Hank Nussbacher hank at att.net.il
Mon Nov 24 10:30:22 EST 2003


On Mon, 24 Nov 2003, Lucas Iglesias wrote:

No passive way to do it?  -Hank

> To properly graph packet loss over a link, you would need to send packets
> (ICMP) secuencially over the link and check they return.
> The packet loss clould happen for 4 reasons (in the roundtrip): local output
> drops, remote input errors, remote output drops and local input errors.
> The only monitoring system that I know for sure it does that is the JFFNMS
> (http://jffnms.sourceforge.net/) ,but I think there has to be others.
> 
> regards,
> Luckas.
> 
> -----Mensaje original-----
> De: Roldan, Brad [mailto:BRoldan at covad.com]
> Enviado el: Viernes, 21 de Noviembre de 2003 09:00 p.m.
> Para: cisco-nsp at puck.nether.net
> Asunto: RE: [nsp] Graphing packet loss
> 
> 
> 
> 
>  I would suggest also including the "ignored" counters in your
> calculation. These provide you with a generally good indication of
> packets getting lost on your VIPs due to congestion. The MIB object is
> 'LocIfIgnored' for most Cisco platforms, or 'CieIfInIgnored' for newer
> IOS releases.
> 
> Brad
> 
> -----Original Message-----
> From: cisco-nsp-bounces at puck.nether.net
> [mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Hank Nussbacher
> Sent: Friday, November 21, 2003 1:34 AM
> To: cisco-nsp at puck.nether.net
> Subject: [nsp] Graphing packet loss
> 
> 
> We would like to graph the percentage pkt loss per line via MRTG polling
> of Cisco routers (7200s and 7500s).I would think that
> (ifOutDisards+ifOutErrors)/ifOutUcastPkts
> (ifInDiscards+ifInErrors)/ifInUcastpkts
> should give the needed result.But others are not sure.  We run
> policy-map shaping on some of the lines and some of the class maps rate
> limit p2p things like Kazaa and Gnutella.That would cause an
> OutDiscard
> probably.So perhaps adding in Discards is not wise?  But if an FE is
> rate limited to 45Mb/sec and the shaping is causing discards, those I
> would want to see.
> 
> And what could cause an iFInDiscard?
> 
> Or is there some better algorithm (specific MIBs plz) people use for
> calculating pkt loss?
> 
> Thanks,
> -Hank
> 
> _______________________________________________
> cisco-nsp mailing listcisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
> 
> _______________________________________________
> cisco-nsp mailing listcisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
> _______________________________________________
> cisco-nsp mailing listcisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
> 

Hank Nussbacher




More information about the cisco-nsp mailing list