alt.certification.cisco - Which has higher priority, administrative distance or longest prefix match

Author

Which has higher priority, administrative distance or longest prefix match

Jason Thomas

2002-06-24, 6:40 pm

Another question, if you all don't mind.

Can anyone tell me which has higher priority administrative distance or
longest prefix match.

Meaning is EIGRP knows of a route to a host and matches 22 bits and OSPF
knows a route and matches 24 bits would the router pick the EIGRP route or
the OSPF route.

I just need a bit of clarification

Thanks,
Jason

karate-kid

2002-06-24, 6:40 pm

Are you sure Hansang? I thought the longest prefix match is always used
since the route would be the most specific to that destination - regardless
of admin distance ? I bow to your greater experience....
Steve

"Hansang Bae" <hbae_@_nyc.rr.com.REMOVE_> wrote in message
news:MPG.17806fd36beef9c7989b92@news-server.nyc.rr.com...
> In article <vSsR8.78541$Na1.2573488@news4.srv.hcvlny.cv.net>,
> jthomas@optonline.net says...
> > Another question, if you all don't mind.
> > Can anyone tell me which has higher priority administrative distance or
> > longest prefix match.
> > Meaning is EIGRP knows of a route to a host and matches 22 bits and OSPF
> > knows a route and matches 24 bits would the router pick the EIGRP route
or
> > the OSPF route.
> > I just need a bit of clarification
>
>
> Administrative distance determines which routes will end up in the routing
> table. OSPF route will not be in the table if EIGRP has the same route
> (regardless of the mask)
>
>
> --
>
> hsb
>
> "Somehow I imagined this experience would be more rewarding" Calvin
> ******************************
******************************
********
> Due to the volume of email that I receive, I may not not be able to
> reply to emails sent to my account. Please post a followup instead.
> ******************************
******************************
********

Hansang Bae

2002-06-24, 6:40 pm

In article <vSsR8.78541$Na1.2573488@news4.srv.hcvlny.cv.net>,
jthomas@optonline.net says...
> Another question, if you all don't mind.
> Can anyone tell me which has higher priority administrative distance or
> longest prefix match.
> Meaning is EIGRP knows of a route to a host and matches 22 bits and OSPF
> knows a route and matches 24 bits would the router pick the EIGRP route or
> the OSPF route.
> I just need a bit of clarification

Administrative distance determines which routes will end up in the routing
table. OSPF route will not be in the table if EIGRP has the same route
(regardless of the mask)

--

hsb

"Somehow I imagined this experience would be more rewarding" Calvin
******************************
******************************
********
Due to the volume of email that I receive, I may not not be able to
reply to emails sent to my account. Please post a followup instead.
******************************
******************************
********

Thomas Kuborn

2002-06-24, 6:40 pm

Yooo,

If you have 2 routes (matching the same # bits) in your routing table,
then AD is first used as a differentiator, then metric is used.
1. # bits matched
2. AD
3. metric
In your example, the OSPF route will be used

Tom

doan wrote:

> I think the longest match has higher priority. You have to think of
> routing as a two step process: building the routing table and selecting
> a route in the routing table. In building the routing table, a router
> will select a route with the best administrative distance, e.g. OSPF
> is better than RIP. Since different prefix-lengths are considered as
> different routes, they will all be installed in the routing table
> regardless of what routing protocol (ad. distance) they were learned
> from. In selecting a route from the routing table, the longest match
> is the rule, administrative distance is no longer an issue here.
>
> Doan
>
> On Sun, 23 Jun 2002, Jason Thomas wrote:
>
>
>>Another question, if you all don't mind.
>>
>>Can anyone tell me which has higher priority administrative distance or
>>longest prefix match.
>>
>>Meaning is EIGRP knows of a route to a host and matches 22 bits and OSPF
>>knows a route and matches 24 bits would the router pick the EIGRP route or
>>the OSPF route.
>>
>>I just need a bit of clarification
>>
>>Thanks,
>>Jason
>>
>>
>>
>>
>

Hansang Bae

2002-06-24, 6:40 pm

In article <Pine.GSO.4.33.0206232221110.17769-100000@talitha.usc.edu>,
doan@talitha.usc.edu says...
> But if the mask is different, then it is a different route, i.e.
> 192.168.32.0/24 and 192.168.32.0/28 are two routes.

Yes they would. The more specific one will be in the routing table. But
this is really a hack and can lead to troubleshooting headaches. While I've
used longer-match routes as a temporary fixes, one shouldn't depend on them
as a general rule of thumb.

--

hsb

"Somehow I imagined this experience would be more rewarding" Calvin
******************************
******************************
********
Due to the volume of email that I receive, I may not not be able to
reply to emails sent to my account. Please post a followup instead.
******************************
******************************
********

Thomas Kuborn

2002-06-24, 6:40 pm

You're right. Sorry

Tom

"doan" <doan@talitha.usc.edu> wrote in message
news:Pine.GSO.4.33.0206232224340.17769-100000@talitha.usc.edu...
>
> I don't think you can have 2 same routes (same # of bits) in the routing
> table.
>
> Doan
>
> On Mon, 24 Jun 2002, Thomas Kuborn wrote:
>
> > Yooo,
> >
> > If you have 2 routes (matching the same # bits) in your routing table,
> > then AD is first used as a differentiator, then metric is used.
> > 1. # bits matched
> > 2. AD
> > 3. metric
> > In your example, the OSPF route will be used
> >
> > Tom
> >
> > doan wrote:
> >
> > > I think the longest match has higher priority. You have to think of
> > > routing as a two step process: building the routing table and
selecting
> > > a route in the routing table. In building the routing table, a router
> > > will select a route with the best administrative distance, e.g. OSPF
> > > is better than RIP. Since different prefix-lengths are considered as
> > > different routes, they will all be installed in the routing table
> > > regardless of what routing protocol (ad. distance) they were learned
> > > from. In selecting a route from the routing table, the longest match
> > > is the rule, administrative distance is no longer an issue here.
> > >
> > > Doan
> > >
> > > On Sun, 23 Jun 2002, Jason Thomas wrote:
> > >
> > >
> > >>Another question, if you all don't mind.
> > >>
> > >>Can anyone tell me which has higher priority administrative distance
or[co
lor=darkred]
> > >>longest prefix match.
> > >>
> > >>Meaning is EIGRP knows of a route to a host and matches 22 bits and[/color]
OSPF[
color=darkred]
> > >>knows a route and matches 24 bits would the router pick the EIGRP[/color]
route or[co
lor=darkred]
> > >>the OSPF route.
> > >>
> > >>I just need a bit of clarification
> > >>
> > >>Thanks,
> > >>Jason
> > >>
> > >>
> > >>
> > >>
> > >
> >
> >
>[/color]

karate-kid

2002-06-24, 6:40 pm

karate-kid

2002-06-24, 6:40 pm

Hansang Bae

2002-06-24, 8:25 pm

In article <xiIR8.2830$k_6.96145@newsfep1-win.server.ntli.net>,
kirkster@ntlworld.com says...
> Are you sure Hansang? I thought the longest prefix match is always used
> since the route would be the most specific to that destination - regardless
> of admin distance ? I bow to your greater experience....

Sorry. See my other post.

--

hsb

"Somehow I imagined this experience would be more rewarding" Calvin
******************************
******************************
********
Due to the volume of email that I receive, I may not not be able to
reply to emails sent to my account. Please post a followup instead.
******************************
******************************
********

Karl Hungus

2002-06-24, 8:25 pm

I think the hansang charater is right, the AD is the biggest "worry"
between reditribution b/w IGRP and OSPF, no OSPF route will ever over
write the IGRP routes, because the OPSF has the AD of 110, and IGRP is
100. When you reditribute these it is required that you adjust the AD
if you want the OSPF routen in your table.

On Mon, 24 Jun 2002 18:04:40 +0100, "karate-kid"
<kirkster@ntlworld.com> wrote:

>Are you sure Hansang? I thought the longest prefix match is always used
>since the route would be the most specific to that destination - regardless
>of admin distance ? I bow to your greater experience....
>Steve
>
>"Hansang Bae" <hbae_@_nyc.rr.com.REMOVE_> wrote in message
>news:MPG.17806fd36beef9c7989b92@news-server.nyc.rr.com...
>> In article <vSsR8.78541$Na1.2573488@news4.srv.hcvlny.cv.net>,
>> jthomas@optonline.net says...
>> > Another question, if you all don't mind.
>> > Can anyone tell me which has higher priority administrative distance or
>> > longest prefix match.
>> > Meaning is EIGRP knows of a route to a host and matches 22 bits and OSPF
>> > knows a route and matches 24 bits would the router pick the EIGRP route
> or

>> > the OSPF route.
>> > I just need a bit of clarification
>>
>>
>> Administrative distance determines which routes will end up in the routing
>> table. OSPF route will not be in the table if EIGRP has the same route
>> (regardless of the mask)
>>
>>
>> --
>>
>> hsb
>>
>> "Somehow I imagined this experience would be more rewarding" Calvin
>> ******************************
******************************
********
>> Due to the volume of email that I receive, I may not not be able to
>> reply to emails sent to my account. Please post a followup instead.
>> ******************************
******************************
********
>

Michael T. Hall

2002-06-26, 7:25 am

"Hansang Bae" <hbae_@_nyc.rr.com.REMOVE_> wrote in message
news:MPG.178081feb1872249989b95@news-server.nyc.rr.com...
> In article <Pine.GSO.4.33.0206232221110.17769-100000@talitha.usc.edu>,
> doan@talitha.usc.edu says...
> > But if the mask is different, then it is a different route, i.e.
> > 192.168.32.0/24 and 192.168.32.0/28 are two routes.
>
>
> Yes they would. The more specific one will be in the routing table. But
> this is really a hack and can lead to troubleshooting headaches. While
I've
> used longer-match routes as a temporary fixes, one shouldn't depend on
them
> as a general rule of thumb.
>

Your comment is ridiculous. There are quite a few reasons why someone would
desire multiple routing entries with varying prefix lengths in their tables.
One common practice is to place static routes to the "null0" interface in
order to populate the routing table with a specific network, in order to
advertise the route via BGP. Then more-specific routes, generated by IGPs,
are used to determine proper interior switching paths.

Yes. Many good practices exercise dependencies upon multiple route entries
with varying prefixes. In all cases, the closest matching prefix (longest
mask) is used to determine a route.

--
Michael T. Hall
michaelthall@attbi.com

Hansang Bae

2002-06-27, 9:25 pm

In article <F3jS8.189179$6m5.159062@rwcrnsc51.ops.asp.att.net>,
michaelthall@attbi.com says...
> Your comment is ridiculous. There are quite a few reasons why someone would
> desire multiple routing entries with varying prefix lengths in their tables.
> One common practice is to place static routes to the "null0" interface in
> order to populate the routing table with a specific network, in order to
> advertise the route via BGP. Then more-specific routes, generated by IGPs,
> are used to determine proper interior switching paths.
> Yes. Many good practices exercise dependencies upon multiple route entries
> with varying prefixes. In all cases, the closest matching prefix (longest
> mask) is used to determine a route.

I should have explained it some background. I'm talking about people who
take the time to summarize networks (in OSPF) for a particular area only to
hijack a small subnet in a different area. Again, a bad idea.

--

hsb

"Somehow I imagined this experience would be more rewarding" Calvin
******************************
******************************
********
Due to the volume of email that I receive, I may not not be able to
reply to emails sent to my account. Please post a followup instead.
******************************
******************************
********

Sponsored Links