[mrtg] Re: Cisco PortChannels

[Nabil Fares]

Hello James,

When you configure EtherChannel/PortChannel, you're actually getting
400mbps.  EtherChannel works on full duplex technology, if you tie 2 fast
Ethernet ports you're getting 2X200mbps total.  When you configure
ethernetchannel on Cisco, it gives both interfaces/ports the same mac
address and one ip address.  I'm not sure how mrtg deals with this.  I would
add
--ifref=ip to make sure you're monitoring one interface.  I hope this
helped!


I almost forgot the essential point.  Ethechannel does load balancing/load
sharing  on TCP/IP protocol only.  IPX and other protocols use only the
first channel.

Nabil

-----Original Message-----
From: mrtg-bounce at list.ee.ethz.ch [mailto:mrtg-bounce at list.ee.ethz.ch]On
Behalf Of Grendel
Sent: Thursday, February 08, 2001 2:35 AM
To: mrtg at list.ee.ethz.ch
Subject: [mrtg] Cisco PortChannels



Hello,
	We have MRTG 2.8.9 running on Solaris, monitoring a Cisco 7513 that's
acting as one of our border routers with the rest of the world. Two of
the FastEthernet ports are tied together into a PortChannel and
connected to one of the Foundry BigIrons at the core of our network.
	Last week there was a hiccup somewhere in the outside world that caused
more than 100 Mbps to (try to?) flow out through this port channel for
the first time. It averages around 25 Mbps, usually peaks only at about
40 Mbps, and with the port channel tie-up should have a max of 200 Mbps,
or at least that's what we all believe here.
	What happened to the graph was a surprise, and I wish I had saved the
daily graph at the time. The "Out" blue line rose to around 50 at about
the time of the external trouble and then plunged to about 300 Kbps and
hovered there for about an hour. Meanwhile the "In" green graph rose to
about 75 and hovered there during the same period. Then as the In graph
started to drop, the Out graph suddenly shot up to just under 100 Mbps
and plateaued there for about an hour.
	Since then I did the math and figured out that MRTG using SNMPv1 and
the default 5 minute interval can graph a max of 114 Mbps. This would
explain the region of the graph that hovered at a measly 300 Kbps. But
it doesn't explain the ceiling it hit at almost 100 Mbps on the other
side of that region.

	Is there someone with experience monitoring or using Cisco port
channels who could explain what I'm seeing? When you tie two ports
together in a port channel on a Cisco and a 've' interface on a Foundry
BigIron at the other end, do you in fact get a load-balanced, cumulative
bandwidth of 200 Mbps, or just a load balanced 100 Mbps?

TIA

James Overbeck
interQ
Tokyo, Japan

--
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