From: Fred Ingham (fningham@xxxxxxxxxxxxxxxx)
Date: Fri Aug 06 1999 - 01:39:28 GMT-3
I think your problem is with your addressing. The addresses break out
as follows:
172.17.59.0
r1 s0.1, r4 s0.1 172.17.59.0 area 1 to r4 /29 248
172.17.59.15 OSPF/IGRP
r1 s0.2, r2 s0 172.17.59.16 area 0 to r2, r3 /28 240
r3 s0 172.17.59.31 OSPF/IGRP
r1 s1, r5 s0 172.17.59.64 s1 to r5 /26 192
172.17.59.127 IGRP
172.17.59.128 r2, r3 To0 /28 240
172.17.59.143 area 3 OSPF
172.17.59.160 r4 e0 /30 252
172.17.59.163 area 2 OSPF
172.17.59.192 r1 e0 /29 248
172.17.59.199 IGRP
The problem is that you should summarize the 128 and 160 subnets on r2,
r3, and r4 with area range commands but you do not have enough address
space
to do so without an overlap if you use a 26 bit mask. The summary
address
command is used to summarize external nets into OSPF. The reason it
worked in IOS 11.2 (my guess) is that the OSPF routes were redistributed
into IGRP and then redistributed back into OSPF where they were
summarized and the summarized routes
were redistributed back to IGRP.
Did you choose this address space or was it given to you?
There is little problem in spacing the addresses for summarization if
the IGRP
serial is configured with a 27 or 28 bit mask. You might try the 26 bit
summarization on r2, r3, and r4 to see if OSPF bauks at the overlap.
Otherwise, change the IGRP serial mask and space out the addresses.
Another way is to use the ip default network command. This will insert
a default
network in IGRP for addresses not summarized.
Hope this helps.
Mark Mirrotto wrote:
>
> All -
>
> I am still having the same ospf summary address problem after I changed the
> network statements. Attached are the configs. Other people that share this
> lab equipment are having the same issues - only sometimes. It started
> happening after r1 was upgraded to 12.0 code.
>
> Thanks for you input to this problem. The workaround is to either add a
> loopback in the summarized range, or change the routers in the 128 subnet
> to something else.
>
> Mark
> -----Original Message-----
> From: Fred Ingham <fningham@worldnet.att.net>
> To: Mark Mirrotto <mmirrott@stratos.net>
> Cc: ccielab@groupstudy.com <ccielab@groupstudy.com>
> Date: Sunday, August 01, 1999 8:17 AM
> Subject: Re: OSPF summary address problem
>
> >I don't know what you're missing but I could not recreate your problem.
> >I configured the setup you described, used the three summary addresses
> >you described, and had the three summary routes in the IGRP domain. No
> >static routes. All routers could ping all interfaces. Attached are the
> >router configurations and the routing tables. Let me know the
> >difference between your configurations and the attached.
> >
> >Another way to insert a default route into IGRP is with the ip
> >default-network command.
> >
> >
> >Mark Mirrotto wrote:
> >>
> >> Hi
> >>
> >> I apologize for the length of this message, but I think it is necessary
> to
> >> set up the scenario.
> >>
> >> R1, R2, and R3 are in a frame relay point-to-multipoint network in subnet
> >> 172.17.59.16 / 28 in OSPF area 0
> >> R3 R2 share a token ring segment using subnet 172.17.59.128 / 28 in OSPF
> >> area 3
> >> R1 to R4 are connected via frame-relay point to point and use subnet
> >> 172.17.59.0 / 28 in OSPF area 1
> >> R4's ethernet segment is in subnet 172.17.59.160 / 30 in OSPF area 2
> >> R1's ethernet is in subnet 172.17.59.192 / 29
> >> R1 to R5 is a standard serial link and is running IGRP only in subnet
> >> 172.17.59.64 / 26
> >> I know I need to summarize because of the classful nature of IGRP, so I
> >> summarize 172.17.59.0 / 26 ; 172.17.59.128 / 26 and 172.17.59.192 / 26 on
> R1
> >> and redistribute igrp and ospf mutually - (I used metrics for
> >> igrp and subnets for ospf) When I do a 'show ip ospf summary' the subnets
> >> all the proper show up summarized, but the 128 subnet has a very high
> >> metric, and doesn't^Òt show up on R5. I think this is because R1 's
> route to
> >> the 128 subnet is a O IA route. I create a static route on R1 with the
> 128
> >> subnet and a 26 bit mask pointing
> >> to null 0 and everything works fine. What am I missing? I won't be able
> to
> >> use a static route in the lab....
> >>
> >> Thanks!
> >>
> >> Mark (31 days and counting.....)
> >>
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 08:21:46 GMT-3