>> On 29 May 2018, at 18:53, Erik Sundberg <[email protected]> wrote: >> I ran into this bug going from 6.3.1 to 6.3.2.... Very simple fix by >> applying to smu patch files. >> https://bst.cloudapps.cisco.com/bugsearch/bug/CSCvf01652?emailclick=CNSemail >> >> Not sure when this change in behavior happened, but by default BGP doesn't >> install/download/show VPNv4 routes unless the VRF is built on the system. >> You need to configure 'retain route-target all' under the vpnv4 address >> family >> >> router bgp 12345 >> address-family vpnv4 unicast >> retain route-target all
On 29 May 2018 at 18:58, Ted Pelas Johansson <[email protected]> wrote: > Hi Erik, > > That is the default of XR. Yeah that was my opinion too. If you are using RRs and the PEs/ASBRs retained all routes reflected to them by default, even when the VRF isn't present and/or "retain route-target xxx" isn't configured, it would cause some scaling issues, so this is exactly how I would expect it to work (and how I have been using IOS-XR). For a PE device we add the VRF to terminate a customer and thus the routes are retained. On an ASBR where we have Inter-AS OptB we use "retain route-target policy xxx" to avoid having all the VRFs defined on a device without the customers connect to it. Kind regards, James. _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
