Hey Marc,
To emphasise, this is for the 33.33.33.33 loopback, or Cat3 loopback route.

I see:
R2-R5-Cat3
R2-R6-Cat4-Cat3
R2-R4-Cat1-Cat3

Why would i go through Cat2 ? So in my count it should prefer the route through 
R5?
On Jul 12, 2011, at 11:05 PM, marc abel wrote:

> You can see they all  have an equal cost [110/66] of 66. Looking at
> your diagram I see
> 
> R2-R4-Cat1-Cat2-Cat2 loopback
> R2-R5-Cat3-Cat2-Cat2 loopback
> R2-R6-Cat4-Cat2-Cat2 loopback
> 
> Extra fiddling is one of the best ways to learn this stuff.
> 
> On Tue, Jul 12, 2011 at 4:49 PM, Alef <[email protected]> wrote:
>> Thanks Marc.
>> 
>> This was the problem:
>> R6#sh class-map type inspect
>>  Class Map type inspect match-any corporate-to-private-class (id 1)
>>   Match protocol telnet
>>   Match protocol ssh
>>   Match protocol ftp
>> 
>> Once i added icmp all was well. I didn't actually expect the route to go 
>> over that for some reason, nor do i understand why it had r6 in it's path 
>> with equal distance, as there is at least one more l3 switch in between 
>> compared to R5. Same for R4. Does it not see the "switches" as a hop or 
>> something ?
>> O       33.33.33.33 [110/66] via 100.100.100.6, 00:13:27, Serial1/1.256
>>                    [110/66] via 100.100.100.5, 00:13:27, Serial1/1.256
>>                    [110/66] via 100.100.24.4, 00:07:08, Serial1/0.24
>> 
>> This is what i get for fiddling with labs designed for a different purpose 
>> and then trying to add more functionality (or get more out of it) ;-)
>> 
>> 
>> On Jul 12, 2011, at 10:36 PM, marc abel wrote:
>> 
>>> Layer 2 loops are caused because switches by default flood traffic out
>>> all ports, so any time you connect switches in a circular path a loop
>>> would form if not for spanning-tree.
>>> 
>>> In layer 3 this is not the case. Since you have to specifically tell
>>> traffic where to route, making a circle does not make a loop. Routes
>>> make loops by referring back to a host that thinks its path is back to
>>> itself. So adding layer 3 interfaces is not the cause of the loops,
>>> incorrect routes are.
>>> 
>>> Take this example
>>> 
>>> R1.
>>> int g0/0
>>> ip address 10.10.10.1 255.255.255.0
>>> no shut
>>> !
>>> ip route  6.6.6.6 255.255.255.255 10.10.10.2
>>> 
>>> R2.
>>> int g0/0
>>> ip address 10.10.10.2 255.255.255.0
>>> no shut
>>> !
>>> ip route  6.6.6.6 255.255.255.255 10.10.10.1
>>> 
>>> 
>>> Now try to traceroute to 6.6.6.6. R1 will send it to R2, R2 will send
>>> it to R1, R1 will send to R2, repeat until the TTL expires. This is a
>>> layer 3 loop. You will generally notice layer 3 loops by traffic not
>>> reaching its destination(and proved by traceroute), or by routes being
>>> repeatedly installed and removed from the routing table (debug ip
>>> routing).
>>> 
>>> Layer 2 loops will generally either keep looping the traffic around so
>>> much that you lose access to the device and the whole damn network
>>> crashes, or if you are lucky you might not have it so bad and you will
>>> see a message in the logs that says something like host x.x.x.x is
>>> flapping between interface.....
>>> 
>>> -Marc
>>> 
>>> 
>>> 
>>> 
>>> On Tue, Jul 12, 2011 at 3:37 PM, Alef <[email protected]> wrote:
>>>> No no redistribution. I tried to keep it as simple as possible.
>>>> Ok let me try my hand at some ascii art.
>>>> 
>>>> But before i do that, is there some inherent stupid thing about putting 
>>>> vlan interfaces in all the vlans on all the switches ( i have dot1q trunks 
>>>> in x between all 4 switches). I can imagine in my network it would create 
>>>> a loop. Just a hunch if i follow the paths. Would i not automatically 
>>>> introduce L3 loops by doing that. I think ipexpert topology is setup 
>>>> similar ?
>>>> 
>>>> ok scratch that, i uploaded a small jpg
>>>> http://tinypic.com/r/fkqfqr/7
>>>> 
>>>> All switches loopbacks are reachable by at least 3 paths (2 trough the FR 
>>>> cloud, and 1 other)
>>>> 
>>>> R2#
>>>> O       11.11.11.11 [110/66] via 100.100.100.6, 01:01:51, Serial1/1.256
>>>>                    [110/66] via 100.100.100.5, 01:01:51, Serial1/1.256
>>>>                    [110/66] via 100.100.24.4, 00:00:43, Serial1/0.24
>>>>     44.0.0.0/32 is subnetted, 1 subnets
>>>> O       44.44.44.44 [110/66] via 100.100.100.6, 01:01:51, Serial1/1.256
>>>>                    [110/66] via 100.100.100.5, 01:01:51, Serial1/1.256
>>>>                    [110/66] via 100.100.24.4, 00:00:43, Serial1/0.24
>>>> 
>>>> On a probably unrelated side note, anyone ever got :
>>>> %OSPF-4-ERRRCV: Received invalid packet: Bad Checksum from 100.100.14.1, 
>>>> Vlan14
>>>> 
>>>> Cisco website says change the port or cable, which i did, and cable, well 
>>>> considering it's a dynamips sim i did  that as best as i could :-)
>>>> 
>>>> Can't get the message to go away though.
>>>> 
>>>> On Jul 12, 2011, at 5:24 PM, Di Bias, Steve wrote:
>>>> 
>>>>> 
>>>>> What makes you think you have any kind of loop? What are the symptoms? 
>>>>> Are you redistributing anywhere?
>>>>> 
>>>>> -----Original Message-----
>>>>> From: [email protected] 
>>>>> [mailto:[email protected]] On Behalf Of Alef
>>>>> Sent: Tuesday, July 12, 2011 8:27 AM
>>>>> To: [email protected] IE
>>>>> Subject: [OSL | CCIE_RS] loop in network?
>>>>> 
>>>>> All,
>>>>> 
>>>>> So i have this network, and i am thinking i have a loop. I'm just not 
>>>>> sure where or how.
>>>>> And i think it started happening after i created vlan interfaces in all 
>>>>> vlans on all switches. So Cat1 is having vlan14,vlan567 and vlan69 
>>>>> interfaces and all the other switches in similar.
>>>>> 
>>>>> They are NM16ESW cards, but IEEE stp is running so a Layer2 loop should 
>>>>> be out of the question. Could i still have a routing loop?
>>>>> I am running OSPF in one giant area 0.
>>>>> 
>>>>> How would i go about solving it? I can't use split horizon as OSPF 
>>>>> doesn't use that. Should i artificially set routes such that they get a 
>>>>> lower metric and manipulate things like that?
>>>>> 
>>>>> I attached a small picture to clarify (this is taken from the Security 
>>>>> lab Day4 VoD btw).
>>>>> Mind you, it might be something different entirely although i don't think 
>>>>> so.
>>>>> 
>>>>> But it's probably going to be something very stupid. And yes, i probably 
>>>>> don't understand switching very well yet :-)
>>>>> 
>>>>> Alef
>>>>> 
>>>>> 
>>>>> UHS Confidentiality Notice:  This e-mail message, including any 
>>>>> attachments, is for the sole use of the intended recipient (s) and may 
>>>>> contain confidential and privileged information.  Any unauthorized 
>>>>> review, use, disclosure or distribution of this information is 
>>>>> prohibited.  If this was sent to you in error, please notify the sender 
>>>>> by reply e-mail and destroy all copies of the original message.
>>>> 
>>>> _______________________________________________
>>>> For more information regarding industry leading CCIE Lab training, please 
>>>> visit www.ipexpert.com
>>>> 
>>>> Are you a CCNP or CCIE and looking for a job? Check out 
>>>> www.PlatinumPlacement.com
>>>> 
>> 
>> 

_______________________________________________
For more information regarding industry leading CCIE Lab training, please visit 
www.ipexpert.com

Are you a CCNP or CCIE and looking for a job? Check out 
www.PlatinumPlacement.com

Reply via email to