From: ying chang (ying_c@xxxxxxxxxxx)
Date: Mon May 06 2002 - 14:44:12 GMT-3
Ahhhh. Picture is the king of all languages. You'll want to have 2 virtual
links r3-r1 and r3-r2, this will allow both r1 and r2 to reach A0 regardless
whether you cut r3-r1 or r3-r2 link. You don't need to build virutal link
between r1 and r2.
Chang
>From: steven owen <trueccie@yahoo.com>
>To: Sean C <Upp_and_Upp@hotmail.com>, ying chang <ying_c@hotmail.com>,
>tron@huapi.ba.ar
>CC: ccielab@groupstudy.com
>Subject: Re: ospf virtual link
>Date: Mon, 6 May 2002 09:56:19 -0700 (PDT)
>
>Sorry for answering so late .
>The scenario is like this,
>
> R3---A0
> > > > > |
> > > > > FR >
> / \
> > > > > R2 R1
> > > > > | |
> ----------
> > > > > ETHERNET
> > > > >
>R1,R2,R3's FR ints are on area 1,R1,R2's E ints are on
>area 2,R3 has an int on area 0.
>First i should build a vl between R1 and R3 ,or R2 and
>R3 to link area 2 with area 0.One vl is enough for
>area 2?
>Do i need to build a vl between R1 and R2?
>Now it seems unnecesary.
>
>Thanks.
>
>
>
>
>
>
>
>--- Sean C <Upp_and_Upp@hotmail.com> wrote:
> > Hi Chang,
> >
> > Who knows?? You're interpretation is just as valid
> > or as crazy as mine.
> > And you're right on the discontiguous stuff - no-go
> > the way I'm reading the
> > topology. Steven hasn't replied to multiple posts -
> > puitting this post to
> > bed.
> >
> > Don't be concerned on the way you read English - at
> > least you have that
> > skill!! I have a hard enough time just deciphering
> > ISDN output - let alone
> > trying a 2nd language. ;-)
> >
> > Sean
> > ----- Original Message -----
> > From: "ying chang" <ying_c@hotmail.com>
> > To: <Upp_and_Upp@hotmail.com>; <tron@huapi.ba.ar>;
> > <trueccie@yahoo.com>
> > Cc: <ccielab@groupstudy.com>
> > Sent: Monday, May 06, 2002 10:20 AM
> > Subject: Re: ospf virtual link
> >
> >
> > > Hi Sean,
> > >
> > > When it comes to interprt the requirements, I'm in
> > deep trouble. Yes.
> > other
> > > than they both are in area 2, Steven did not say
> > they are connected or
> > not.
> > > If they are not connected, aren't we getting
> > ourselves into this
> > > discontiguous area 2 stuff - two area 2s seperate
> > by area 0 and area 1.
> > If
> > > this is the case, there's no phyical backup in the
> > topology anywhere,
> > until
> > > we get into Harry Potter's magic stuff, I agree
> > with you, we'll loose
> > > connectivity if we loose a link.
> > >
> > > You know, you read English from left to right, I
> > read Chinese from top to
> > > down, seeing from the remote, kind of like you
> > shaking your head and
> > saying
> > > "no, no, no..." and I'm noding my head saying
> > "yes, yes, yes...", can this
> > > be why I mistakenly connect put two links
> > together?
> > >
> > > Thanks,
> > > Chang
> > >
> > >
> > > >From: "Sean C" <Upp_and_Upp@hotmail.com>
> > > >To: "ying chang" <ying_c@hotmail.com>,
> > <tron@huapi.ba.ar>,
> > > ><trueccie@yahoo.com>
> > > >CC: <ccielab@groupstudy.com>
> > > >Subject: Re: ospf virtual link
> > > >Date: Mon, 6 May 2002 09:25:22 -0400
> > > >
> > > >Hi Chang,
> > > >
> > > >The way I was reading Steven's topology, I didn't
> > see any link between r2
> > > >and r3 besides the FR connection on Area 1.
> > > >"r1,r2,r3 connected by FR,all the interfaces on
> > FR are on area 1"
> > > >
> > > >Your map has a Eth link between R2 and R3 - I
> > read the description as an
> > > >Eth
> > > >link between R2 and R1.
> > > >"r1 and r2 have an ethernet int each, which is on
> > area 2, r3 have an int
> > on
> > > >area 0."
> > > >
> > > >If your interpretation of Steven's description is
> > correct, then yes, the
> > > >virtual-link will work. If the only link to R3
> > is by the frame, and R3's
> > > >frame fails - then they are out of luck.
> > > >
> > > >Am I reading this incorrectly?
> > > >Sean
> > > >
> > > >
> > > >
> > > >----- Original Message -----
> > > >From: "ying chang" <ying_c@hotmail.com>
> > > >To: <tron@huapi.ba.ar>; <trueccie@yahoo.com>
> > > >Cc: <ccielab@groupstudy.com>
> > > >Sent: Monday, May 06, 2002 8:16 AM
> > > >Subject: Re: ospf virtual link
> > > >
> > > >
> > > > > If I understand Steven's topology correctly, I
> > cannot see why not. The
> > > >way
> > > >I
> > > > > see his topology is like below:
> > > > >
> > > > > R1
> > > > > / \
> > > > > A1 A1
> > > > > / \
> > > > > R2 R3---A0
> > > > > | |
> > > > > |--A2----|
> > > > >
> > > > > A1 is in the Frame Relay cloud, A2 and A0 are
> > in LAN interfaces, if
> > > >R1-R3
> > > >FR
> > > > > link fails, he wants be able to reach A0 via
> > R1-R2-R3 link. If this is
> > > >the
> > > > > case, I see it's a perfect case for virtual
> > link between R1 and R3
> > just
> > > >by
> > > > > cutting the link betwwen R1-R3 link:
> > > > >
> > > > > R2:
> > > > >
> > > > > router ospf 100
> > > > > ...
> > > > > area 2 virtual-link R3's RID
> > > > > ...
> > > > >
> > > > > R3:
> > > > >
> > > > > router ospf 100
> > > > > ...
> > > > > area 2 virtual-link R2's RID
> > > > > ...
> > > > >
> > > > > Since the error message is referring to RID
> > mismatch, I would use
> > "show
> > > >ip
> > > > > ospf" to find both R2 and R3's RID. I'm
> > guessing he probably peering
> > R1
> > > >and
> > > > > R3 instead of R2 and R3, but I'm not sure.
> > > > >
> > > > > Actually, if you want to get fancy, you can
> > use A1 to backup A2 as
> > well.
> > > > > i.e. A2 is the primary route for R2 to reach
> > A0, but if A2 fails, R2
> > > >should
> > > > > be able to use A1 to reach A0.
> > > > >
> > > > > Chang
> > > > >
> > > > >
> > > > > >From: Carlos G Mendioroz <tron@huapi.ba.ar>
> > > > > >Reply-To: Carlos G Mendioroz
> > <tron@huapi.ba.ar>
> > > > > >To: steven owen <trueccie@yahoo.com>
> > > > > >CC: Groupstudy ccielab list
> > <ccielab@groupstudy.com>
> > > > > >Subject: Re: ospf virtual link
> > > > > >Date: Mon, 06 May 2002 07:52:00 -0300
> > > > > >
> > > > > >Steven,
> > > > > >virtual links are always area 0.
> > > > > >They are use to make up for the need of area
> > 0 being THE
> > > > > >connecting area (aka backbone) and in one
> > piece.
> > > > > >Tipical uses have one end in real area 0, and
> > the other in
> > > > > >a remote (i.e. non directly connected) area,
> > but it also
> > > > > >can be used to reattach a disconnected area 0
> > sector
> > > > > >(thus both ends lying in area 0).
> > > > > >
> > > > > >Now, answering you question, no, you can not.
> > > > > >(Also you need not, since it is ok for having
> > more than
> > > > > >one area X at one time.)
> > > > > >What you do need is having area 2 directly
> > connected to
> > > > > >area 0, and there you can (should) use one or
> > two
> > > > > >VLs to go via area 1.
> > > > > >
> > > > > >steven owen wrote:
> > > > > > >
> >
>=== message truncated ===
>
>
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 10:58:51 GMT-3