Interesting what draws comment and what doesn't. Howard's puzzle is quite a
challenge, when one approaches it with reason rather than with routers.
In general I was 2/3's right ( arguably ) but I missed a major point when I
stated my predictions.
---------------
Howard's scenario:
Scenario 1 (R1 is initially misconfigured)
R1 E0: 10.6.0.1 DOWN R2 E0: 10.1.0.1 UP/UP
E1: 10.5.0.2 UP/UP E1: 10.5.0.2 UP/UP
E2: 10.2.0.1 UP/UP E2: 10.2.0.2 UP/UP
Admin discovers that R1 E1 is misconfigured and should have been 10.5.0.1.
Our Heroine corrects that interface to 10.5.0.1.
Assuming both routers had OSPF configured with
network 10.0.0.0 0.255.255.255 area 0.0.0.1
Will a device on R1 E2 be able to ping a host on R2 E2?
--------------
My feeble attempt at cleverness: ( and actual results noted )
[ I had to use loopbacks to simulate the multiple ethernet ports ]
>Assuming that there is no typo in your addressing scheme, I say you have a
>real mess here and that the answer is no.
CL: this much was true. But this was a case of me being right, but for the
wrong reason. See below.
>1) you have the same subnet appearing on two different routers
>10.0.5.0/whatever on R1E1 and R2E1 While OSPF can handle discontiguous
>subnets, I would be surprised if it can handle duplicate subnets with any
>aplomb
CL: OSPF definitely did NOT like duplicate networks. Name a routing protocol
that would?
>2) on R1 the interface with the address of 10.0.5.2 has been placed into
>area 0.0.0.1. when the ip address for that interface is changed, without
>either reloading the router or performing a clear ip ospf process the
>interface with the address 10.5.0.1 will NOT have been placed into the OSPF
>process.
CL: absolutely WRONG.
CL: What really happens is that the routers in question never form adjacency
because they both have the same router i.d. RID's must be unique. With the
duplicate RID's, the routers do not form neighbor relationships. The LSA's
aren't processed. A DR and a BDR are not formed. No routing information is
propagated. End of story
CL: my thanks to Kevin W. for his observation, which caused me to see the
error of my initial Q&D scenario. The key piece I overlooked was the
interface with the address of 10.6.0.1, which should have been the RID on
router one. However, with the interface down, it was not. The next highest
address - 10.5.0.2 became the RID. Trouble begins.
CL: BTW, even after correcting the interface IP address, the problem
continues. So, then, I have a further question for all of you. How do you
correct this problem? Or rather, what are some things one might suppose
would correct this problem. Does that work? Is there a "clear" command one
can use? Or must one resort to blowing away the OSPF process, and then
restoring it?
Hhhmmmmm........ does this qualify as a Friday Folly? :->
Chuck
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of
Chuck Larrieu
Sent: Wednesday, November 29, 2000 9:47 PM
To: Chuck Larrieu; Howard C. Berkowitz; [EMAIL PROTECTED]
Subject: RE: About OSPF and Loopback port
OK, I have run a couple of Q&D's. I have some observations, but I will save
them until a few of you have taken a stab at Howard's puzzle for yourselves.
No fair touching routers first. You gotta test your knowledge and your
reasoning.
I'll share my results Thursday evening.
Chuck
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of
Chuck Larrieu
Sent: Wednesday, November 29, 2000 8:47 PM
To: Howard C. Berkowitz; [EMAIL PROTECTED]
Subject: RE: About OSPF and Loopback port
OK Howard. I'm not afraid to look foolish in front of everyone.
-----------------------
Howard's scenario:
Scenario 1 (R1 is initially misconfigured)
R1 E0: 10.6.0.1 DOWN R2 E0: 10.1.0.1 UP/UP
E1: 10.5.0.2 UP/UP E1: 10.5.0.2 UP/UP
E2: 10.2.0.1 UP/UP E2: 10.2.0.2 UP/UP
Admin discovers that R1 E1 is misconfigured and should have been 10.5.0.1.
Our Heroine corrects that interface to 10.5.0.1.
Assuming both routers had OSPF configured with
network 10.0.0.0 0.255.255.255 area 0.0.0.1
Will a device on R1 E2 be able to ping a host on R2 E2?
----------------------
My feeble attempt at cleverness:
Assuming that there is no typo in your addressing scheme, I say you have a
real mess here and that the answer is no.
1) you have the same subnet appearing on two different routers
10.0.5.0/whatever on R1E1 and R2E1 While OSPF can handle discontiguous
subnets, I would be surprised if it can handle duplicate subnets with any
aplomb
2) on R1 the interface with the address of 10.0.5.2 has been placed into
area 0.0.0.1. when the ip address for that interface is changed, without
either reloading the router or performing a clear ip ospf process the
interface with the address 10.5.0.1 will NOT have been placed into the OSPF
process.
3) I have not done this on my routers yet. I want to see if I am in the
right ballpark so far. ( well, to be honest, as soon as I send this message
I am setting up a scenario, upon which I will report subsequently )
Well, gang - am I wise or still a fool?
Chuck
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of
Howard C. Berkowitz
Sent: Wednesday, November 29, 2000 12:37 PM
To: [EMAIL PROTECTED]
Subject: RE: About OSPF and Loopback port
At 10:45 AM -0800 11/29/2000, Healis, Jim wrote:
>Use loopback ports in OSPF so you can set the Router OSPF ID, otherwise it
>will take the highest IP address.
>
>Jim Healis CCNP, CCDP
>Senior Network Administrator
>Virata
Unless there have been recent IOS changes (I'm really most current in
11-something), it's even more unpredictable than just the highest IP
address:
At the time of OSPF initialization, the router ID is:
if there are multiple loopback interfaces, the highest IP address on any
loopback (i.e., not highest loopback interface number)
if there is a single loopback interface, use its address
if there are no loopback interfaces, use the highest IP address on any
active interface (i.e., if all interfaces are in shutdown, OSPF
can't initialize. Using loopbacks avoids this because a loopback
cannot be down.)
A fiendish troubleshooting scenario:
R1 comes up first, then R2. They share an Ethernet. Neither has
any loopbacks.
Scenario 1 (R1 is initially misconfigured)
R1 E0: 10.6.0.1 DOWN R2 E0: 10.1.0.1 UP/UP
E1: 10.5.0.2 UP/UP E1: 10.5.0.2 UP/UP
E2: 10.2.0.1 UP/UP E2: 10.2.0.2 UP/UP
Admin discovers that R1 E1 is misconfigured and should have been 10.5.0.1.
Our Heroine corrects that interface to 10.5.0.1.
Assuming both routers had OSPF configured with
network 10.0.0.0 0.255.255.255 area 0.0.0.1
Will a device on R1 E2 be able to ping a host on R2 E2?
>
>
> -----Original Message-----
>From: Moerdo [mailto:[EMAIL PROTECTED]]
>Sent: Wednesday, November 29, 2000 8:26 AM
>To: [EMAIL PROTECTED]
>Subject: About OSPF and Loopback port
>
>Does anyone here can explain to me, why me must use loopback port for OSPF
>configuration. Thank you for the answer for this stupid question. Thank
you.
>
>moerdo.
>
_________________________________
FAQ, list archives, and subscription info:
http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
_________________________________
FAQ, list archives, and subscription info:
http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
_________________________________
FAQ, list archives, and subscription info:
http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
_________________________________
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
