OSPF Network Point-to-Multipoint

Just to let you know,

As per one of the explanations:

R8 does not have an exact match for the prefix 155.1.0.0/24, which is the actual
subnet assignment of the DMVPN Network. Instead, it knows that there are five
endpoints on the network: 155.1.0.1, 155.1.0.2, 155.1.0.3, 155.1.0.4, and 155.1.0.5.
This is the normal and desirable behavior for OSPF network type point-to-multipoint,
according to the RFC standard.

On CSR 1000v IOS XE 3.11S the route is learnt on R8.  This is a BUG (another one). 

I tried this on my real rack and GNS3 and the /24 route of the DMVPN is not advertised at all by R5 to R8.

This is the 2nd weird thing I've seen on this IOS XE version, so I'm upgrading to 3.1202S in hope to see less of this.  Anyways I'm doing the whole V5 WB again so I'll try to post as many bugs as I can whenever I found one.

 

Comments

  • Yup it will change the next-hop value, it must.  Recall that OSPF consider these as a collection of Point-to-Point towards R5. On broadcast net, the value is supposed to be maintained.

    Regarding the leaking Route, I upgraded yesterday to XE 3.12.02. The /24 route is not there anymore, so yes it was a bug:

    R8#sh ip route 155.1.0.0 255.255.255.0
    % Subnet not in table

    R8#sh ip route ospf | i 155.1.0.
          155.1.0.0/16 is variably subnetted, 11 subnets, 2 masks
    O IA     155.1.0.1/32 [110/1001] via 155.1.58.5, 00:06:09, GigabitEthernet1.58 <--- /32's
    O IA     155.1.0.2/32 [110/1001] via 155.1.58.5, 00:06:09, GigabitEthernet1.58 <--- /32's
    O IA     155.1.0.3/32 [110/1001] via 155.1.58.5, 00:06:09, GigabitEthernet1.58 <--- /32's
    O IA     155.1.0.4/32 [110/1001] via 155.1.58.5, 00:06:09, GigabitEthernet1.58 <--- /32's
    O IA     155.1.0.5/32 [110/1] via 155.1.58.5, 00:06:09, GigabitEthernet1.58 <--- /32's

    I hope this IOS version to be better. Whenever I start to see strange behaviours on the routers I tend to panic a little bit, were supposed to waste time
    troubleshooting the labs not the IOS...man it "BUGS" me!!! [:P]

    Anyways I'm giving the CSRs a try since is what INE uses now.  But let me tell you GNS3 3725's are ridiculous stable. It is true that it lacks of some new features like EIGRP Named Mode, but you still can do allmost everything there, flawless.

  • If you use CSRs, why did you use 3.11 instead of 3.13?

     

  • If you use CSRs, why did you use 3.11 instead of 3.13?

     

    Oh, that's because that was the more stable version I got working on Wmware Workstation 10.  3.12.00 was crashing all the time.

    Then I got this "new one" 3.12.02 which is stable. 

    The main reason I'm not using the 3.13 is because the licensing, for me 100k of throughput is too little.  In order to activate de Eval license, I would need

    to request one per virtual device, contrary to the 3.12 and lower. On those the only thing you need is to run the "license boot level premium" and reboot. And

    the other thing I noitced is the 60days reset every time I power down/up the appliances, which is great...I guess.

    Anyways what's your experience with 3.13 ? It is better? And how about the throughput capping?  Maybe I'm missing something.

  • BTW,

     

    My virtual gear is running on Linux Mint Qiana.  Vmware Workstation 10.

    AMD 64FX 8- cores | 32G RAM.

    After all the routers are up and running, they consume around 37% CPU and 2.x Gigs of RAM counting the RAM used by the Host OS.

    The only thing I recommed to the fellas who download the OVA, delete the Virtual HDD and create a new one but IDE not SCSI. Delete 1 Network Card and after the IOS is installed remove the CD ROMs as well. Is less HW to virtualize.

    If you happend to have the ISO , just create a VM with 1 or 2 NICs, IDE disk, Cdrom, 2.5GRam and 1CPU.
    This is very minimalistic, yet it works without mayor issues.

     

  • Anyways what's your experience with 3.13 ? It is better? And how about the throughput capping?  Maybe I'm missing something.

    Mate, it's running fine for me (I'm only up to EIGRP and OSPF anyway) so the capping has not bothered me. 

    First problem I have is the route-map (or leak-map) with CSR. Pretty sure I read it somewhere here. Just trying to find it.

  • Thanks for sharing.

    Yeh, I think regardless of the version the bugs and "weird behaviours" are inevitable.  Sometimes I have to switch between CSR, GNS3, IOU, etc.

    At the end of the day nothing better than the real gear, but I know in terms of cost and lack of resources these tools really come in handy.

  • I see your point and I agree. 

    But I wrote "nothing better than the real gear".

    What I meant is that these virtual tools are not yet to perform a 100% and many times you end troubleshooting the IOS rather than the LAB topics.  Having a real rack overcomes many issues related to the virtualization which is more prone to bugs. (In general terms, not only focusing on the exam)

    It is true that the main goal is to pass the exam but knowing when a feature should work is vital as well. That way you'll know when to restart a node during the exam rather than troubleshoot it further.

    But again, I agree and no one could said this better:

     

    Trick is to get same IOU versions as they have in the lab. 

         This is great advice.

Sign In or Register to comment.