Re: Delay Metric in EIGRP [7:36001]

2002-03-09 Thread Hans PHAM
Steven A. Ridder wrote: I believe the delay is by default set on the interface by the router based on the type of link it is. I'm sure there's charts on CCO somewhere. You can change this info on the interface with the delay command, which is the recommended way of changing a metric if

Re: Delay Metric in EIGRP [7:36001]

2002-02-21 Thread Steven A. Ridder
So is it safe for me to conclude that I was wrong in stating that EIGRP sends the metrics to its neighbors. It actually sends the raw data such as bandwidth, and delay, and the neighbor router uses the DUAL FSM process to calculate the Reported Distance and then it's distance? I've been

Re: Delay Metric in EIGRP [7:36001]

2002-02-21 Thread Phil Barker
The 'Bandwidth' and 'Delay' parameters ARE metrics, which are sent to neighbours in packets in order that DUAL can be run to calculate its topological database and routing table. Phil. --- Steven A. Ridder wrote: So is it safe for me to conclude that I was wrong in stating that EIGRP

Re: Delay Metric in EIGRP [7:36001]

2002-02-21 Thread Priscilla Oppenheimer
A router sends the components of the metric. The recipient calculates the actual composite metric by working a bunch of black magic. The DUAL process and figuring out successors are separate from the metric calculation, but use the metric info. Here's a terrific paper on the topic:

Re: Delay Metric in EIGRP [7:36001]

2002-02-21 Thread Priscilla Oppenheimer
Did anyone notice that I explained reliability and load backwards? ;-) It was a copy-and-paste error. Reverse the two explanations below please. Priscilla At 08:43 PM 2/20/02, Priscilla Oppenheimer wrote: The delay part of the EIGRP composite metric is not measured. It's based on the type of

Re: Delay Metric in EIGRP [7:36001]

2002-02-21 Thread Patrick Ramsey
ahhh you ole cut'n'paster :p :) Priscilla Oppenheimer 02/21/02 02:51PM Did anyone notice that I explained reliability and load backwards? ;-) It was a copy-and-paste error. Reverse the two explanations below please. Priscilla At 08:43 PM 2/20/02, Priscilla Oppenheimer wrote: The

Re: Delay Metric in EIGRP [7:36001]

2002-02-21 Thread Priscilla Oppenheimer
The EIGRP bandwidth metric is 2.56 billion divided by the real bandwidth. That's just how Cisco does it (and how Sniffer displays it). In this case, the bandwidth metric is 1657856, as the Sniffer reports. The network is reachable via a single, ordinary 1544 Kbps serial interface. Do the math.

Re: Delay Metric in EIGRP [7:36001]

2002-02-21 Thread Priscilla Oppenheimer
Reposting. I never saw this one get posted. The EIGRP bandwidth metric is 2.56 billion divided by the real bandwidth. That's just how Cisco does it (and how Sniffer displays it). In this case, the bandwidth metric is 1657856, as the Sniffer reports. The network is reachable via a single,

Delay Metric in EIGRP [7:36001]

2002-02-20 Thread Yatou Wu
Hi, In EIGRP, the delay metric is taken as configured in the interface of the router by the administrator, by default, or by measurement? when the router calculates the metric, it needs to know the minimum bandwidth along the path, and also the delay along the path. how can the router pass

Re: Delay Metric in EIGRP [7:36001]

2002-02-20 Thread Steven A. Ridder
I believe the delay is by default set on the interface by the router based on the type of link it is. I'm sure there's charts on CCO somewhere. You can change this info on the interface with the delay command, which is the recommended way of changing a metric if you are forced to do so. The

Re: Delay Metric in EIGRP [7:36001]

2002-02-20 Thread Sasa Milic
The router DOES pass total delay and minumum bandwidth of the route to neighbors. check show ip eigrp topologu Sasa CCIE No 8635 Steven A. Ridder wrote: I believe the delay is by default set on the interface by the router based on the type of link it is. I'm sure there's charts on CCO

Re: Delay Metric in EIGRP [7:36001]

2002-02-20 Thread Steven A. Ridder
It's not in a packet that gets sent. Sasa Milic wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... The router DOES pass total delay and minumum bandwidth of the route to neighbors. check show ip eigrp topologu Sasa CCIE No 8635 Steven A. Ridder wrote: I believe the

Re: Delay Metric in EIGRP [7:36001]

2002-02-20 Thread Steven A. Ridder
Oops, I looked it up and what do you know... these's a nice field for bandwidth and delay. Steven A. Ridder wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... It's not in a packet that gets sent. Sasa Milic wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... The

Re: Delay Metric in EIGRP [7:36001]

2002-02-20 Thread Priscilla Oppenheimer
At 07:03 PM 2/20/02, Steven A. Ridder wrote: It's not in a packet that gets sent. It is actually. Here's a packet for you: DLC: - DLC Header - DLC: Destination = Station Cisco1053E80 DLC: Source = Station Cisco1002E75 DLC: Ethertype = 0800 (IP) IP: -

Re: Delay Metric in EIGRP [7:36001]

2002-02-20 Thread Priscilla Oppenheimer
The delay part of the EIGRP composite metric is not measured. It's based on the type of interface. Each type of interface has a default value. You can change it, although this is risky and not recommended. Because EIGRP is a distance-vector protocol, the router sends route updates that list

Re: Delay Metric in EIGRP [7:36001]

2002-02-20 Thread Chuck
I'm curious about the EIGRP bandwidth being reported at 2.5 billion bps. what kind of a network you running there, PO? ;- see below: Priscilla Oppenheimer wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... At 07:03 PM 2/20/02, Steven A. Ridder wrote: It's not in a packet that

Re: Delay Metric in EIGRP [7:36001]

2002-02-20 Thread Chuck
well, phrak, the server ate my notations: trying again: see below. Priscilla Oppenheimer wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... At 07:03 PM 2/20/02, Steven A. Ridder wrote: It's not in a packet that gets sent. It is actually. Here's a packet for you: DLC: