From: Anthony Pace (anthonypace@xxxxxxxxxxx)
Date: Thu Jun 20 2002 - 01:26:18 GMT-3
I have another question in vein of this thread. What is the distinction
of a "demand-peer"? Aren't they all demand peers if you create a
hierarchy of border peers and groups who have an indirect connection to
all the other routers in all other groups. Do you need to "denote" this
any where or is it "autoconfigured" whenever you create this kind of
hierarchy?
Anthony Pace
On Mon, 17 Jun 2002 12:35:21 -0700 (PDT), "Paul" <p_chopin@yahoo.com>
said:
> I've just tested passive command in my lab.Cisco IOS
> allows you to put this statement either on local
> peer-id or remote peer-id.In case you put on local
> peer-id
> then router won't iniciate any connections to
> configred
> remote peers.It does work.
> Paul
> --- George Spahl <g.spahl@insightbb.com> wrote:
> > Charles,
> > I hope by continuing this thread that we will draw
> > out a DLSW guru to
> > add his or her comments or perhaps correct what I
> > have said.
> > I believe that by adding the passive statement to
> > the local peer command
> > (I think that's the only place it can be added) you
> > prevent that router
> > from initiating any connections to its configured
> > remote peers. So it
> > seems like it is an all or nothing proposition. If
> > you had a one
> > hundred remote peer statements you wouldn't be able
> > to initiate a
> > connection to any of them if you add the passive
> > keyword to your local
> > peer statement. Seems like it would have made more
> > sense for the
> > passive keyword to be added on to the remote-peer
> > statement. That way
> > you could initiate connections selectively. At
> > least this is how I
> > think it works but would like to hear from others
> > who know better.
> > George
> >
> >
> > -----Original Message-----
> > From: nobody@groupstudy.com
> > [mailto:nobody@groupstudy.com] On Behalf Of
> > Carley, Charles
> > Sent: Monday, June 17, 2002 8:27 AM
> > To: 'Michael Popovich'; Paul; elping
> > Cc: CCIE GROUPSTUDY
> > Subject: RE: Dlsw backup peer
> >
> > I see the use of the "promiscuous" command
> > documented well, what is the
> > difference with the "passive" command used in the
> > same context.
> >
> > -----Original Message-----
> > From: Michael Popovich [mailto:m.popovich@mchsi.com]
> >
> > Sent: Monday, June 17, 2002 12:48 AM
> > To: Paul; elping
> > Cc: CCIE GROUPSTUDY
> > Subject: Re: Dlsw backup peer
> >
> > Paul-
> >
> > I agree with you.
> >
> > Here is a link on Cisco:
> >
> http://www.cisco.com/warp/public/697/dlsw_redundancy.html#solution
> >
> > Notice on Solution #3 where backup-peers is
> > discussed. Every solution I
> > have
> > seen with backup-peers and in my lab for testing
> > shows that if your
> > peers
> > are in promiscuous mode then backup-peer works,
> > primary shows connected
> > and
> > backup show disconnected. If the all DLSW routers
> > have defined peers
> > then
> > this is not true.
> >
> > I have not tested in a lab to see if by chance the
> > backup funtionality
> > still
> > works though. I have been wondering and I plan on
> > testing it this week.
> > I am
> > wondering if all DLSW router peers are defined and
> > you still have
> > backup-peer configured if the circuits would
> > function the same.
> >
> > R2------R3
> > | |
> > R4 |
> > |-------Host
> >
> > All routers have peers defined. R2 would show both
> > R3 and R4 in Connect
> > state. If R3 was primary and R4 was backup. Would
> > hosts build circuits
> > through R3 and if R3 lost connectivity to R2 would
> > those cirucuits get
> > torn
> > down in R2 with the ability to rebuild through R4. I
> > would guess yes,
> > but
> > what I would be interested to see, is if R3 came
> > back online any new
> > sessions would establish through R3. I am doubting
> > it. I think "cost" is
> > what should be used with all peers are defined and
> > backup-peer should be
> > used on promiscuous setups.
> >
> > MP
> > ----- Original Message -----
> > From: "Paul" <p_chopin@yahoo.com>
> > To: "elping" <elpingu@acedsl.com>
> > Cc: <ccielab@groupstudy.com>
> > Sent: Sunday, June 16, 2002 11:07 PM
> > Subject: Re: Dlsw backup peer
> >
> >
> > > I read somewhere that if we have all peers
> > configured
> > > with dlsw remote statements(instead of
> > promiscuous)
> > > then dlsw backup peer will be overwritten and the
> > > state will show up as connect.I wonder what is
> > > solution in this case. What's gonna happened if
> > > primary peer (R2) has higher cost than backup (R4)
> > > --- elping <elpingu@acedsl.com> wrote:
> > > > no.
> > > > backup peers will be disconnected till the
> > primary
> > > > is down.
> > > >
> > > > Paul wrote:
> > > >
> > > > > Hi group,
> > > > > I wonder if it is normal for backup dlsw peer
> > to
> > > > have
> > > > > state connect.I always assumed that backup
> > peer
> > > > kicks
> > > > > in when primary connection goes down.
> > > > > I have R2 and R4 routers attached to the same
> > > > token
> > > > > ring. R1 primary session supposed to be to R2
> > and
> > > > in
> > > > > case R2 is down , R1 should peer to R4.I have
> > > > > configured backup peer with linger command on
> > R1
> > > > ,but
> > > > > the connection to R4 stays up all the time.
> > > > > All routers have remote statements hardcoded.
> > We
> > > > are
> > > > > not allowed to use promiscious mode on anyone
> > of
> > > > the
> > > > > routers.
> > > > > Did anybody run to the same problem?
> > > > > What am I doing wrong? Should I use border
> > group
> > > > > peers?
> > > > > Thanks.Paul
> > > > > Here are simple configs:
> > > > > R1
> > > > > dlsw local-peer peer-id 139.1.1.1
> > > > > dlsw remote-peer 0 tcp 139.1.2.2
> > > > > dlsw remote-peer 0 tcp 139.1.4.4 backup-peer
> > > > 139.1.2.2
> > > > > linger 5
> > > > >
> > > > > R2
> > > > > dlsw local-peer peer-id 139.1.2.2
> > > > > dlsw remote-peer 0 tcp 139.1.1.1
> > > > >
> > > > > R4
> > > > > dlsw local-peer peer-id 139.1.4.4 cost 2
> > > > > dlsw remote-peer 0 tcp 139.1.1.1
> > > > >
> > > > >
This archive was generated by hypermail 2.1.4 : Tue Jul 02 2002 - 08:12:38 GMT-3