Redistribute static under router bgp zzzzz vrf xxxxx does not work

[DanielGutierrez615]

I was working on a maintenance window.  moving some customers to our  new MPLS network  PE router NCS540 SOFTWARE 7.6.2.

When I move 192.168.70.0/24 to the new mpls network   I was able to ping it from everywhere. however a static routed network I was not able to reach even though I could see the route advertized everywhere in the network

int bvi 100

vrf CORE_14288

ipv4 address 192.168.70.1 255.255.255.0

!

router bgp 11749
vrf CORE_14288
rd auto
address-family ipv4 unicast
redistribute connected
redistribute static
!

!

router static
vrf CORE_14288
address-family ipv4 unicast

192.168.10.0/27 192.168.70.2

After doing some digging I found that the redistribute static was not  allowing the traffic to pass thru even though I could see it advertized from every PE in the network.

If I hardcode it like shown below, it works. 

router bgp xxxx

vrf CORE_14288
address-family ipv4 unicast
192.168.10.0/27 192.168.70.2
!
!

I opened a ticket with cisco just to make sure this is not a bug, but I'm curious to see if anybody has experience the same behaviour. 

[DanielGutierrez615]

I was following this article

NCS5500 Routing in VRF Cisco NCS5500 @xrdocs

[MHM Cisco World]

workaround for this case you can try
assign static label for static route 
and keep per-ce mode 

MHM

[Harold Ritter]

Hi @MHM Cisco World ,

The issue is related to the per CE label allocation mode behavior, which is LFIB lookup and forward directly to adjacency. Assigning a static label will not solve the issue, as behavior will be the same but with static label instead. @DanielGutierrez615 has already tried it unsuccessfully, as mentioned previously.

Regards,

[MHM Cisco World]

but making PE assign label statically prevent it from assign two label, 
we see that other PE see two label for same prefix

MHM

[Harold Ritter]

Hi @MHM Cisco World ,

The issue is not with the two labels. It is due to traffic using the per CE label, which does not seem to work as it should.

Regards,

[DanielGutierrez615]

Are there any study cases on this specific topic Label Allocation modes you’re aware OF. I want to make sure I understand this 100%.I have like 50 tabs open on my browser trying to find more information about it, but all of them pretty much say the same thing. I’m looking for something that will dig deeper into the subject  

[MHM Cisco World]

https://www.cisco.com/c/en/us/support/docs/multiprotocol-label-switching-mpls/multiprotocol-label-switching-mpls/217583-configure-vrf-label-mode-with-carrier-su.html

Open this link now you will have 51 tab lol..

Check this link it good 

MHM

[Harold Ritter]

Hi @DanielGutierrez615 ,

As mentioned in the paper you referred to, in doubts use per CE (which is the default anyhow on the device you are using). That is why you need to continue working with TAC to find out if the behavior you are seeing is a bug or a limitation with this specific hardware and configuration combination.

Regards,

[DanielGutierrez615]

will do.   I will let you guys know what info I get from them . Thank you guys for your feedback.  you helped me a ton on this tshoot case.

[nkarpysh]

Hi @DanielGutierrez615 ,

You need to use per-vrf Label mode over BVI. It does not support any other Label modes. You can configure it under BGP VRF.

[Ronit Bhattacharjee]

We ran across the same problem today with our NCS540s. With per-prefix label distribution, cannot reach networks advertised by redistributing static routes. Changing to per-CE works well.

Redistributing connected routes works fine in both cases.

We have opened a case with Cisco to get to the bottom of this. 

[smilstea]

Per-prefix mode is in general a bad idea, especially on a platform like NCS540.
You are probably running into a scale problem but I don't have enough information.
It is recommended on NCS540 to not use per-prefix. There are a few documents that discuss this and even in this thread.

Thanks,
Sam

[Ronit Bhattacharjee]

Thanks, it is unlikely to be a scale issue, because we have less than 50 prefixes in our lab network where we are testing this. Either ways, we have a case with TAC to work on this further,

[Harold Ritter]

Hi @Ronit Bhattacharjee ,

This is a known limitation.

https://www.cisco.com/c/en/us/td/docs/iosxr/ncs5500/interfaces/79x/b-interfaces-hardware-component-cg-ncs5500-79x/configuring-integrated-routing-bridging.html

"Label allocation mode per-CE with BVI is not supported in an access network along with PE-CE protocols enabled"

Regards,

[Ronit Bhattacharjee]

Thanks, I did see this, but it doesn't match our issue exactly

1. There's no issue with the BVI subnet's prefix being advertised across the network. All BVI-BVI routing works without issues with "redistribute connected". The problem is only with "redistribute static".
2. Label-allocation mode "per-CE" is the only mode that actually works for us. The problem is "per-Prefix"