Re: Was Re: longest match vs. other metrics [7:41692]
>Related question to the above thread. > >As discussed, each routing protocol will maintain its list of prefixes >that it knows about. A route selection process runs that considers the >routes from each routing process and puts the "best" into the routing >table. (best being defined by the route selection process). > >"show ip route" will show those routes selected and in the routing >table. > >Is there an equivalent command that will show me the same information >(prefix/length, next hop/interface) for all the prefixs known by a >routing protocol? Suppose I'm running OSPF and BGP on a box and I >wanted to see what prefix/lengths are being carried in each routing >protocol. What commands would I use to see this? > >Thanks Well, they don't precisely create routing tables, but show ip ospf database and show ip bgp will come close. Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=41825&t=41692 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
RE: Was Re: longest match vs. other metrics [7:41692]
show ip route ospf show ip bgp show ip route ? -Original Message- From: bergenpeak [mailto:[EMAIL PROTECTED]] Sent: 18 April 2002 13:37 To: [EMAIL PROTECTED] Subject: Was Re: longest match vs. other metrics [7:41692] Related question to the above thread. As discussed, each routing protocol will maintain its list of prefixes that it knows about. A route selection process runs that considers the routes from each routing process and puts the "best" into the routing table. (best being defined by the route selection process). "show ip route" will show those routes selected and in the routing table. Is there an equivalent command that will show me the same information (prefix/length, next hop/interface) for all the prefixs known by a routing protocol? Suppose I'm running OSPF and BGP on a box and I wanted to see what prefix/lengths are being carried in each routing protocol. What commands would I use to see this? Thanks Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=41817&t=41692 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
Was Re: longest match vs. other metrics [7:41692]
Related question to the above thread. As discussed, each routing protocol will maintain its list of prefixes that it knows about. A route selection process runs that considers the routes from each routing process and puts the "best" into the routing table. (best being defined by the route selection process). "show ip route" will show those routes selected and in the routing table. Is there an equivalent command that will show me the same information (prefix/length, next hop/interface) for all the prefixs known by a routing protocol? Suppose I'm running OSPF and BGP on a box and I wanted to see what prefix/lengths are being carried in each routing protocol. What commands would I use to see this? Thanks Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=41816&t=41692 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
Re: longest match vs. other metrics [7:41692]
I would agree with you that it wouldn't load balance over paths with differing ADs. However, theoretically, you could have a dynamic route learned via a routing protocol (having a given AD) and you could add a static route with the same length (i.e. /24) and purposely set the AD to match the one learned from the routing protocol and it would be installed in the routing table along with the dynamic route and marked for load balancing. Anyone ever try this? Or would the metric from the dynamic route come into play there? If the route with best AD gets put in the routing table (among multiple routes with the same length), where does the metric come into play? (Wow... it's been a while since BSCN =) But I would also agree that most of the time, usually load balancing usually occurs over equal cost paths (and implied equal lengths as well) that are learned from the same protocol (and thus have the same AD) Mike W. "Jeff Smith" wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > I am under the impression that if a router had more than one route with the > same prefix length in its database(s) then it would choose the one with the > lowest AD to place in its routing table. As in John's example only when the > prefixes are different are they both in the routing table. AFAIK, a Cisco > router will load balance between paths but only when it involves the same > protocol, it will not load balance between differing ADs. Please educate me > if I am incorrect. > > Jeff Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=41714&t=41692 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
Re: longest match vs. other metrics [7:41692]
I am under the impression that if a router had more than one route with the same prefix length in its database(s) then it would choose the one with the lowest AD to place in its routing table. As in John's example only when the prefixes are different are they both in the routing table. AFAIK, a Cisco router will load balance between paths but only when it involves the same protocol, it will not load balance between differing ADs. Please educate me if I am incorrect. Jeff >From: "Chuck" >Reply-To: "Chuck" >To: [EMAIL PROTECTED] >Subject: Re: longest match vs. other metrics [7:41692] >Date: Tue, 16 Apr 2002 23:05:34 -0400 > >perhaps I am misunderstanding your answer. > >a router always forwards based on the longest match. this is a requirement, >per RFC 1812. > >if there are two or more routes in a routing table of the same prefix >length, then, in accordance with the rules regarding load sharing, more >than >one route / interface can be used. > >admin distance is the tiebreaker a ( Cisco ) router uses when determining >which routes of identical prefix length, but learned from different routing >protocols, to place into the routing table in the first place. > >John N's post on this topic is an excellent explanation of the two >different >processes. > >Chuck > > >""Jeff Smith"" wrote in message >[EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > > Sean, the longest match is the first characteristic considered. So, >using > > your example, IOS will choose the route that has the longest prefix >match- > > only when they are the same will the decison come down to administratice > > distance between protocols. > > > > P.S. > > Due to your connections can I get some Red Sox tickets? > > > > > > >From: "Sean Wolfe" > > >Reply-To: "Sean Wolfe" > > >To: [EMAIL PROTECTED] > > >Subject: longest match vs. other metrics [7:41692] > > >Date: Tue, 16 Apr 2002 22:08:54 -0400 > > > > > >Quick question, hope it's not too trivial: > > > > > >When a router decides to forward a packet based on the longest match > > >principle, does this supersede other factors? > > > > > >For example, if there is a route to network A via EIGRP, but a more > > >specific > > >route available via OSPF, does it choose OSPF because of longest match, >or > > >EIGRP because of lower administrative distance (90 vs. 110)? > > > > > >Thanks folks, fun reading your posts as always. Wish me luck as I take >BCSN > > >this week. -Sean. > > _ > > Chat with friends online, try MSN Messenger: http://messenger.msn.com _ Chat with friends online, try MSN Messenger: http://messenger.msn.com Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=41712&t=41692 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
Re: longest match vs. other metrics [7:41692]
Rules based on RFC1812 with industry extensions: 1. If route is not in table, add it. 2. If route is more specific than one in the table, add it, keeping the old one. 3. If the routes are of equal length, install the one with highest preference (GateD, Juniper, Bay) or lowest AD (Cisco). Replace any route from another source. On Cisco, if the routes are from a static source and have the same AD, add the new route to the table (up to maximum-paths) and mark it eligible for load-sharing. 4. If the routes are of the same AD/preference, and the protocol has a metric, replace the one in the table if the new one has a lower metric. If the metrics are the same, add the route and make it eligible for load-sharing. >I believe its distance to determine the protocol, then the longest >match after that... > >Larry Letterman >Cisco Systems >[EMAIL PROTECTED] >- Original Message - >From: "Sean Wolfe" >To: >Sent: Tuesday, April 16, 2002 7:08 PM >Subject: longest match vs. other metrics [7:41692] > > >> Quick question, hope it's not too trivial: >> >> When a router decides to forward a packet based on the longest match >> principle, does this supersede other factors? >> >> For example, if there is a route to network A via EIGRP, but a more >specific >> route available via OSPF, does it choose OSPF because of longest match, or >> EIGRP because of lower administrative distance (90 vs. 110)? >> >> Thanks folks, fun reading your posts as always. Wish me luck as I take >BCSN >> this week. -Sean. Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=41711&t=41692 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
Re: longest match vs. other metrics [7:41692]
Looks like I should read it, since I was mistaken..:) Larry Letterman Cisco Systems [EMAIL PROTECTED] - Original Message - From: "Chuck" To: Sent: Tuesday, April 16, 2002 8:00 PM Subject: Re: longest match vs. other metrics [7:41692] > excellent explanation, John. > > to plug a source of study materials, Howard Berkowitz has an excellent paper > on how routers work, including the route determination process on > certification zone ( www.certificationzone.com ) > > per RFC 1812, all routers forward based on the longest match. > > Chuck > > > > ""John Neiberger"" wrote in message > [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > > There are two separate processes involved here and it's > > important to make a distinction between them. There is the > > process by which the router builds the routing table and then > > there is the route selection process. > > > > When building the routing table, the router looks at the > > various routes available to it. For any given route, if it's > > learned from multiple protocols the Administrative Distance > > wins. It's important to remember that AD comes into play when > > comparing identical prefixes. > > > > For example, let's say you learn 205.243.23.0/24 via EIGRP and > > via OSPF. The router will install the EIGRP route into the > > routing table because its AD is lower. > > > > However, if the router learns 205.243.23.0/24 via EIGRP and > > 205.243.23.0/25 via OSPF, both routes will be installed because > > the prefix length (subnet mask) is different. > > > > The router does this comparison--if necessary--with all the > > routes it has learned from all available sources and then > > compiles a final routing table. > > > > Now, a packet arrives that needs to be forwarded. It is here > > that the longest match rule really applies. The route that > > most closely matches the destination for the packet is the > > route that will be used. Using our previous example, let's say > > a packet is destined for 205.243.23.42. In this case, the /25 > > route learned via OSPF will be chosen. > > > > If the destination were 205.243.23.150, though, then the /24 > > route would be the closest match. > > > > HTH, > > John > > > > On Tue, 16 Apr 2002, Sean Wolfe > > ([EMAIL PROTECTED]) wrote: > > > > > Quick question, hope it's not too trivial: > > > > > > When a router decides to forward a packet based on the > > longest match > > > principle, does this supersede other factors? > > > > > > For example, if there is a route to network A via EIGRP, but > > a more specific > > > route available via OSPF, does it choose OSPF because of > > longest match, or > > > EIGRP because of lower administrative distance (90 vs. 110)? > > > > > > Thanks folks, fun reading your posts as always. Wish me luck > > as I take BCSN > > > this week. -Sean. > > [EMAIL PROTECTED] Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=41710&t=41692 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
RE: longest match vs. other metrics [7:41692]
The router will always take the more specific route regardless of administrative distance. Below is the routing table of a router running EIGRP and OSPF, as you can see 55.55.0.0/16 is from EIGRP, but the router still takes the 55.55.55.0/24 route from OSPF because it is more specific. No question is too trivial, believe me I have asked some winners in my time :), good luck on your test. r6#sh ip route Codes: C - connected, S - static, I - IGRP, R - RIP, M - mobile, B - BGP D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2 E1 - OSPF external type 1, E2 - OSPF external type 2, E - EGP i - IS-IS, L1 - IS-IS level-1, L2 - IS-IS level-2, * - candidate default U - per-user static route, o - ODR Gateway of last resort is not set 55.0.0.0/8 is variably subnetted, 2 subnets, 2 masks D 55.55.0.0/16 [90/2297856] via 6.6.6.2, 00:00:02, Serial0 O 55.55.55.0/24 [110/74] via 6.6.6.5, 00:36:53, Serial0 5.0.0.0/24 is subnetted, 1 subnets O E25.5.5.0 [110/20] via 6.6.6.5, 00:03:05, Serial0 6.0.0.0/24 is subnetted, 1 subnets C 6.6.6.0 is directly connected, Serial0 r6# -Original Message- From: Sean Wolfe To: [EMAIL PROTECTED] Sent: 4/16/2002 10:08 PM Subject: longest match vs. other metrics [7:41692] Quick question, hope it's not too trivial: When a router decides to forward a packet based on the longest match principle, does this supersede other factors? For example, if there is a route to network A via EIGRP, but a more specific route available via OSPF, does it choose OSPF because of longest match, or EIGRP because of lower administrative distance (90 vs. 110)? Thanks folks, fun reading your posts as always. Wish me luck as I take BCSN this week. -Sean. Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=41707&t=41692 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
Re: longest match vs. other metrics [7:41692]
perhaps I am misunderstanding your answer. a router always forwards based on the longest match. this is a requirement, per RFC 1812. if there are two or more routes in a routing table of the same prefix length, then, in accordance with the rules regarding load sharing, more than one route / interface can be used. admin distance is the tiebreaker a ( Cisco ) router uses when determining which routes of identical prefix length, but learned from different routing protocols, to place into the routing table in the first place. John N's post on this topic is an excellent explanation of the two different processes. Chuck ""Jeff Smith"" wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > Sean, the longest match is the first characteristic considered. So, using > your example, IOS will choose the route that has the longest prefix match- > only when they are the same will the decison come down to administratice > distance between protocols. > > P.S. > Due to your connections can I get some Red Sox tickets? > > > >From: "Sean Wolfe" > >Reply-To: "Sean Wolfe" > >To: [EMAIL PROTECTED] > >Subject: longest match vs. other metrics [7:41692] > >Date: Tue, 16 Apr 2002 22:08:54 -0400 > > > >Quick question, hope it's not too trivial: > > > >When a router decides to forward a packet based on the longest match > >principle, does this supersede other factors? > > > >For example, if there is a route to network A via EIGRP, but a more > >specific > >route available via OSPF, does it choose OSPF because of longest match, or > >EIGRP because of lower administrative distance (90 vs. 110)? > > > >Thanks folks, fun reading your posts as always. Wish me luck as I take BCSN > >this week. -Sean. > _ > Chat with friends online, try MSN Messenger: http://messenger.msn.com Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=41706&t=41692 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
Re: longest match vs. other metrics [7:41692]
excellent explanation, John. to plug a source of study materials, Howard Berkowitz has an excellent paper on how routers work, including the route determination process on certification zone ( www.certificationzone.com ) per RFC 1812, all routers forward based on the longest match. Chuck ""John Neiberger"" wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > There are two separate processes involved here and it's > important to make a distinction between them. There is the > process by which the router builds the routing table and then > there is the route selection process. > > When building the routing table, the router looks at the > various routes available to it. For any given route, if it's > learned from multiple protocols the Administrative Distance > wins. It's important to remember that AD comes into play when > comparing identical prefixes. > > For example, let's say you learn 205.243.23.0/24 via EIGRP and > via OSPF. The router will install the EIGRP route into the > routing table because its AD is lower. > > However, if the router learns 205.243.23.0/24 via EIGRP and > 205.243.23.0/25 via OSPF, both routes will be installed because > the prefix length (subnet mask) is different. > > The router does this comparison--if necessary--with all the > routes it has learned from all available sources and then > compiles a final routing table. > > Now, a packet arrives that needs to be forwarded. It is here > that the longest match rule really applies. The route that > most closely matches the destination for the packet is the > route that will be used. Using our previous example, let's say > a packet is destined for 205.243.23.42. In this case, the /25 > route learned via OSPF will be chosen. > > If the destination were 205.243.23.150, though, then the /24 > route would be the closest match. > > HTH, > John > > On Tue, 16 Apr 2002, Sean Wolfe > ([EMAIL PROTECTED]) wrote: > > > Quick question, hope it's not too trivial: > > > > When a router decides to forward a packet based on the > longest match > > principle, does this supersede other factors? > > > > For example, if there is a route to network A via EIGRP, but > a more specific > > route available via OSPF, does it choose OSPF because of > longest match, or > > EIGRP because of lower administrative distance (90 vs. 110)? > > > > Thanks folks, fun reading your posts as always. Wish me luck > as I take BCSN > > this week. -Sean. > [EMAIL PROTECTED] Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=41704&t=41692 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
Re: longest match vs. other metrics [7:41692]
Sean, the longest match is the first characteristic considered. So, using your example, IOS will choose the route that has the longest prefix match- only when they are the same will the decison come down to administratice distance between protocols. P.S. Due to your connections can I get some Red Sox tickets? >From: "Sean Wolfe" >Reply-To: "Sean Wolfe" >To: [EMAIL PROTECTED] >Subject: longest match vs. other metrics [7:41692] >Date: Tue, 16 Apr 2002 22:08:54 -0400 > >Quick question, hope it's not too trivial: > >When a router decides to forward a packet based on the longest match >principle, does this supersede other factors? > >For example, if there is a route to network A via EIGRP, but a more >specific >route available via OSPF, does it choose OSPF because of longest match, or >EIGRP because of lower administrative distance (90 vs. 110)? > >Thanks folks, fun reading your posts as always. Wish me luck as I take BCSN >this week. -Sean. _ Chat with friends online, try MSN Messenger: http://messenger.msn.com Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=41703&t=41692 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
Re: longest match vs. other metrics [7:41692]
I believe its distance to determine the protocol, then the longest match after that... Larry Letterman Cisco Systems [EMAIL PROTECTED] - Original Message - From: "Sean Wolfe" To: Sent: Tuesday, April 16, 2002 7:08 PM Subject: longest match vs. other metrics [7:41692] > Quick question, hope it's not too trivial: > > When a router decides to forward a packet based on the longest match > principle, does this supersede other factors? > > For example, if there is a route to network A via EIGRP, but a more specific > route available via OSPF, does it choose OSPF because of longest match, or > EIGRP because of lower administrative distance (90 vs. 110)? > > Thanks folks, fun reading your posts as always. Wish me luck as I take BCSN > this week. -Sean. Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=41700&t=41692 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
Re: longest match vs. other metrics [7:41692]
There are two separate processes involved here and it's important to make a distinction between them. There is the process by which the router builds the routing table and then there is the route selection process. When building the routing table, the router looks at the various routes available to it. For any given route, if it's learned from multiple protocols the Administrative Distance wins. It's important to remember that AD comes into play when comparing identical prefixes. For example, let's say you learn 205.243.23.0/24 via EIGRP and via OSPF. The router will install the EIGRP route into the routing table because its AD is lower. However, if the router learns 205.243.23.0/24 via EIGRP and 205.243.23.0/25 via OSPF, both routes will be installed because the prefix length (subnet mask) is different. The router does this comparison--if necessary--with all the routes it has learned from all available sources and then compiles a final routing table. Now, a packet arrives that needs to be forwarded. It is here that the longest match rule really applies. The route that most closely matches the destination for the packet is the route that will be used. Using our previous example, let's say a packet is destined for 205.243.23.42. In this case, the /25 route learned via OSPF will be chosen. If the destination were 205.243.23.150, though, then the /24 route would be the closest match. HTH, John On Tue, 16 Apr 2002, Sean Wolfe ([EMAIL PROTECTED]) wrote: > Quick question, hope it's not too trivial: > > When a router decides to forward a packet based on the longest match > principle, does this supersede other factors? > > For example, if there is a route to network A via EIGRP, but a more specific > route available via OSPF, does it choose OSPF because of longest match, or > EIGRP because of lower administrative distance (90 vs. 110)? > > Thanks folks, fun reading your posts as always. Wish me luck as I take BCSN > this week. -Sean. [EMAIL PROTECTED] Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=41697&t=41692 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
Re: longest match vs. other metrics [7:41692]
i think it choose longest match,
what happend if this case between static route entry and eigrp route entry?
yes ,it choose longest.
but you know static route administrative distance is 1.
""Sean Wolfe"" P4HkO{O"PBNE
:[EMAIL PROTECTED]
> Quick question, hope it's not too trivial:
>
> When a router decides to forward a packet based on the longest match
> principle, does this supersede other factors?
>
> For example, if there is a route to network A via EIGRP, but a more
specific
> route available via OSPF, does it choose OSPF because of longest match, or
> EIGRP because of lower administrative distance (90 vs. 110)?
>
> Thanks folks, fun reading your posts as always. Wish me luck as I take
BCSN
> this week. -Sean.
Message Posted at:
http://www.groupstudy.com/form/read.php?f=7&i=41696&t=41692
--
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
Re: longest match vs. other metrics [7:41692]
If the routes were equal (length), it would only have the eigrp route to chose from, because that's the only route that would be in the table because of the lower AD. The router would still have the route learned from OSPF in it's osopf database, but it wouldn't enter it into the RIB unless it lost the eigrp route. But since the routes aren't equal (length), it will enter both into RIB, and choose longest match, which in your scenario is OSPF. -- RFC 1149 Compliant. Get in my head: http://sar.dynu.com ""Sean Wolfe"" wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > Quick question, hope it's not too trivial: > > When a router decides to forward a packet based on the longest match > principle, does this supersede other factors? > > For example, if there is a route to network A via EIGRP, but a more specific > route available via OSPF, does it choose OSPF because of longest match, or > EIGRP because of lower administrative distance (90 vs. 110)? > > Thanks folks, fun reading your posts as always. Wish me luck as I take BCSN > this week. -Sean. Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=41695&t=41692 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
