Re: [Idr] draft-walton-bgp-hostname-capability-00

[Hannes Gredler <[e-mail]>]

hi daniel,

given the LL nature of the IP addresses (= no DNS) this sounds a reasonable proposal.

assuming that all name information go to some internal database for resolution:
have you put some thought on how to consolidate naming information from different
sources ? (bgp-hostnames, BGP-LS node/link names, LLDP ...)

/hannes

On Sat, May 16, 2015 at 12:08:35AM -0400, Daniel Walton wrote:
|    Many of our customers use a Clos architecture and bgp peer to the IPv6 LL
|    address of every directly connected router.  To avoid configuring neighbor
|    statements with LL addresses we support a "neighbor swpX interface"
|    command ("swpX" is the default notation for an interface name in cumulus
|    linux).  We then peer to the LL address on the other end of the swpX link
|    (we only support this for point-to-point links).
|    When you run commands like "show ip bgp summary" you see a list of
|    interface names instead of IP addresses. Typically if you are loopback
|    peering you either become familiar with "10.0.0.1 is Chicago" because you
|    see 10.0.0.1 in the 'show ip bgp summ' output of several different routers
|    or you could use DNS to do a lookup on 10.0.0.1 and put the hostname in
|    the output.  When every box is peering to LL addresses though ("swpX" for
|    short for us) you can't easily tell who that peer is and you can't use DNS
|    because they LL addresses
|    Before
|    leaf-11# show ip bgp summ
|    BGP router identifier 6.0.0.5, local AS number 65101
|    BGP table version 25
|    RIB entries 33, using 3960 bytes of memory
|    Peers 6, using 100 KiB of memory
|    Peer groups 1, using 56 bytes of memory
|    Neighbor        V    AS MsgRcvd MsgSent   TblVer  InQ OutQ Up/Down
|     State/PfxRcd
|    swp1            4 65000     196     198        0    0    0 00:02:51
|     9
|    swp2            4 65000     195     200        0    0    0 00:02:51
|     9
|    swp3            4 65000     197     200        0    0    0 00:02:51
|     9
|    swp4            4 65000     198     200        0    0    0 00:02:51
|     9
|    swp5            4 65201     207     200        0    0    0 00:02:51
|     2
|    swp6            4 65202     210     201        0    0    0 00:02:52
|     2
|    Total number of neighbors 6
|    leaf-11#
|    spine-3# show ip bgp summ
|    BGP router identifier 6.0.0.15, local AS number 65000
|    BGP table version 19
|    RIB entries 33, using 3960 bytes of memory
|    Peers 8, using 134 KiB of memory
|    Peer groups 1, using 56 bytes of memory
|    Neighbor        V    AS MsgRcvd MsgSent   TblVer  InQ OutQ Up/Down
|     State/PfxRcd
|    swp5            4 65101     675     675        0    0    0 00:10:49
|     5
|    swp6            4 65101     678     676        0    0    0 00:10:50
|     5
|    swp7            4 65101     667     668        0    0    0 00:10:48
|     5
|    swp8            4 65101     676     676        0    0    0 00:10:49
|     5
|    swp9            4 65102     701     692        0    0    0 00:11:04
|     5
|    swp10           4 65102     701     692        0    0    0 00:11:04
|     5
|    swp11           4 65102     701     691        0    0    0 00:11:04
|     5
|    swp12           4 65102     701     692        0    0    0 00:11:04
|     5
|    Total number of neighbors 8
|    spine-3#
|    spine-3#
|    After
|    leaf-11# show ip bgp summ
|    BGP router identifier 6.0.0.5, local AS number 65101
|    BGP table version 25
|    RIB entries 33, using 3960 bytes of memory
|    Peers 6, using 100 KiB of memory
|    Peer groups 1, using 56 bytes of memory
|    Neighbor        V    AS MsgRcvd MsgSent   TblVer  InQ OutQ Up/Down
|     State/PfxRcd
|    spine-1(swp1)   4 65000     211     213        0    0    0 00:03:07
|     9
|    spine-2(swp2)   4 65000     210     215        0    0    0 00:03:07
|     9
|    spine-3(swp3)   4 65000     212     215        0    0    0 00:03:07
|     9
|    spine-4(swp4)   4 65000     213     215        0    0    0 00:03:07
|     9
|    tor-11(swp5)    4 65201     222     215        0    0    0 00:03:07
|     2
|    tor-12(swp6)    4 65202     225     216        0    0    0 00:03:08
|     2
|    Total number of neighbors 6
|    leaf-11#
|    spine-3# show ip bgp summ
|    BGP router identifier 6.0.0.15, local AS number 65000
|    BGP table version 19
|    RIB entries 33, using 3960 bytes of memory
|    Peers 8, using 134 KiB of memory
|    Peer groups 1, using 56 bytes of memory
|    Neighbor        V    AS MsgRcvd MsgSent   TblVer  InQ OutQ Up/Down
|     State/PfxRcd
|    leaf-11(swp5)   4 65101    1119    1119        0    0    0 00:18:13
|     5
|    leaf-12(swp6)   4 65101    1122    1120        0    0    0 00:18:14
|     5
|    leaf-13(swp7)   4 65101    1110    1111        0    0    0 00:18:12
|     5
|    leaf-14(swp8)   4 65101    1119    1119        0    0    0 00:18:13
|     5
|    leaf-21(swp9)   4 65102    1144    1135        0    0    0 00:18:28
|     5
|    leaf-22(swp10)  4 65102    1144    1135        0    0    0 00:18:28
|     5
|    leaf-23(swp11)  4 65102    1144    1134        0    0    0 00:18:28
|     5
|    leaf-24(swp12)  4 65102    1144    1135        0    0    0 00:18:28
|     5
|    Total number of neighbors 8
|    spine-3#
|    It makes troubleshooting much easier.
|    Daniel
|    On Fri, May 15, 2015 at 9:58 PM, <[1][e-mail]> wrote:
| 
|      Agree.. adding code for no real operational benefit
| 
|      Sent from my BlackBerry 10 smartphone on the Rogers network.
|        Original Message
|      From: Randy Bush
|      Sent: Friday, May 15, 2015 10:52 PM
|      To: Daniel Walton
|      Cc: idr wg; Dinesh Dutt
|      Subject: Re: [Idr] draft-walton-bgp-hostname-capability-00
|      one more unneeded place for things to go wrong
| 
|      randy
| 
|      _______________________________________________
|      Idr mailing list
|      [2][e-mail]
|      [3]https://www.ietf.org/mailman/listinfo/idr
| 
| References
| 
|    Visible links
|    1. mailto:[e-mail]
|    2. mailto:[e-mail]
|    3. https://www.ietf.org/mailman/listinfo/idr

| _______________________________________________
| Idr mailing list
| [e-mail]
| https://www.ietf.org/mailman/listinfo/idr