Re: MPLS LDP IGP Sync HOLDDOWN?

So maybe I can restate the question (sorry for the long email).

Cisco config guide says the following regarding LDP IGP Sync feature:

"If the LDP peer is reachable, the IGP waits indefinitely (by default) for
synchronization to be achieved. To limit the length of time the IGP session
must wait, enter the *mpls ldp igp sync holddown *command.

If the LDP peer is not reachable, the IGP establishes the adjacency to
enable the LDP session to be established.

When an IGP adjacency is established on a link but LDP-IGP synchronization
is not yet achieved or is lost, the IGP advertises the max-metric on that
link."

http://www.cisco.com/en/US/customer/docs/ios/mpls/configuration/guide/mp_ldp_igp_synch_ps6922_TSD_Products_Configuration_Guide_Chapter.html#wp1053817

The way I understand this is that LDP IGP Sync will behave differently
depending on the LDP peer transport IP address being reachable.
1. If the LDP peer's transport IP address is reachable - it will keep the
IGP adj state down until LDP neighbor comes up. You can configure a hold
timer to bring it up after it expires as an optional parameter.
2. If the LDP peer's transport IP address IS NOT reachable - it will bring
up the IGP session up, but as long as the LDP peering is still down, the
respective link will be advertised with IGP Max metric. The idea behind
this one is if LDP peer address is not reachable, but discovered, you need
IGP first to establish the LDP session or else LDP won't know how to reach
it.

I hope my interpretation of Cisco config guide is correct. Now, my question
relates to the first rule. I had my router keep the IGP session down even
when the other side wasn't sending LDP Hellos. Without hellos my router
won't know what is the LDP peer address (no discovery) so it can't be
reachable. If it doesn't know what's the LDP peer address, how can it keep
the IGP session down? Is there another rule to this feature that says if a
router DOESN'T know the LDP neighbor's peer address, keep the IGP session
down?

Now when I did some testing and with Diptanshu's help these are the two
rules I can reproduce in lab regarding this feature's rules. Both of these
depend on the order of operations. Is OSPF enabled first or is LDP enabled
first:

If OSPF is up and then you enable LPD (MPLS IP):
--> OSPF will send max metric for the interface until LDP comes up, but it
will keep the IGP session up.

If OSPF is not up or enabled but LDP is configured:
--> OSPF will be kept down until LDP comes up. You can use the hold down to
bring up the OSPF session after timer expires.
--> LDP comes up first, then IGP session is brought up.

(I appreciate the really good discussion on this topic)

Thanks

Tom Kacprzynski

On Thu, Jul 4, 2013 at 1:56 AM, Brian McGahan <bmcgahan_at_ine.com> wrote:

> The feature can only save you so much. If you left "mpls ip" off then it's
> your fault not the feature's fault. The ideal situation is that all links
> run LDP IGP auto config and LDP IGP sync, which by design tries to prevent
> both a config and software problem.
>
> In the MPLS core in a real design all links run IGP/LDP so you would want
> this protection. If a link runs IGP but not LDP you simply turn LDP auto
> config off at the link level for that particular interface.
>
> Brian McGahan, CCIE #8593 (R&S/SP/Security), CCDE #2013::13
> bmcgahan_at_INE.com
>
> Internetwork Expert, Inc.
> http://www.INE.com <http://www.ine.com/>
>
> On Jul 3, 2013, at 11:19 PM, "Tom Kacprzynski" <tom.kac_at_gmail.com> wrote:
>
> Thanks Yuri!!...that's exactly what I was looking for. I could not generate
> an example where the IGP session would stay down. I was only able to get
> the max cost advertised...This brings me to another questions. If the rules
> for LDP IGP Sync say that if the LDP peer is reachable but the LDP session
> is not up, by default the IGP connections will be in down state
> indefinitely (unless you configure the holddown timers). But my question
> is if the LDP session is down how is the router going to know what's the
> LDP peer's transport IP to check if it is reachable? If LDP hellos are
> never sent they will not know about each other. I am missing another rule
> in the LDP IGP Sync feature?
>
> For example I referenced this topology at
>
> http://kemot-net.wdfiles.com/local--files/mpls-ldp-igp-synchronization/mpls%20ldp%20igp%20sync%20001.JPG
>
> R4's Gi2/0 is down, because R3's Gi2/0 does not have "mpls ip" configured.
> IGP and LDP are not synchronized.
>
> R4#sh ip ospf mpls ldp interface gi2/0
> GigabitEthernet2/0
> Process ID 1, Area 0
> LDP is configured through LDP autoconfig
> LDP-IGP Synchronization : Required
> Holddown timer is not configured
> Interface is down and pending LDP <<------------interface down
>
>
> Router R4 does not see the transport IP of the peer, since it doesn't get
> any LDP hellos.
>
> R4#sh mpls ldp discovery detail
> Local LDP Identifier:
> 4.4.4.4:0
> Discovery Sources:
> Interfaces:
> GigabitEthernet1/0 (ldp): xmit/recv
> ....
> GigabitEthernet2/0 (ldp): *xmit*
>
> Enabled: IGP config;
> Hello interval: 5000 ms; Transport IP addr: 4.4.4.4
> Clients: IPv4, mLDP
>
> ...so how is R4 able to detect the reachability of R3's peer IP address per
> the feature's rules?
>
> Thank you,
>
> Tom
>
>
>
>
>
> On Wed, Jul 3, 2013 at 4:25 PM, Yuri Bank <yuribank_at_gmail.com> wrote:
>
> What if you're using LDP autoconfiguration? You may have some links to non
>
> MPLS enabled routers/devices, that are running a IGP (Such as a Load
>
> balancer, firewall etc...) Without specifying a holddown time the IGP
>
> adjacency would never come up. Although in that situation it would be best
>
> to disable mpls entirely on those interfaces.
>
>
> -YuriB
>
>
>
>
>
> On Tue, Jul 2, 2013 at 6:45 PM, Tom Kacprzynski <tom.kac_at_gmail.com> wrote:
>
>
> Hi,
>
> I'm trying to figure out why would anyone use the holddown option for IGP
>
> Sync? Does is not advertise the max metric after the timers expires for
>
> the
>
> link having synchronization issue? If that's the case why would you enable
>
> the sync feature in the first time if it will be bypassed after the
>
> holddown timer expires? Am I missing something?
>
>
> I have a topology I'm testing this feature. Instead of filling up the
>
> email
>
> with attachments I put the image at
>
> http://wiki.kemot-net.com/mpls-ldp-igp-synchronization
>
>
> Thank you,
>
>
> Tom Kacprzynski
>
>
>
> Blogs and organic groups at http://www.ccie.net
>
>
> _______________________________________________________________________
>
> Subscription information may be found at:
>
> http://www.groupstudy.com/list/CCIELab.html
>
>
>
> Blogs and organic groups at http://www.ccie.net
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html

Blogs and organic groups at http://www.ccie.net
Received on Thu Jul 04 2013 - 10:08:37 ART