In general, you are correct. But you have to draw the line SOMEPLACE,
right? :) By default, there can be only one. So in addition:
13. Prefer the path that comes from the lowest neighbor address.
This address is the IP address that is used in the BGP neighbor
configuration. The address corresponds to the remote peer that is used in
the TCP connection with the local router.
Check out:
http://www.cisco.com/warp/public/459/25.shtml
HTH,
Scott Morris, CCIE4 (R&S/ISP-Dial/Security/Service Provider) #4713, JNCIE-M
#153, JNCIS-ER, CISSP, et al.
CCSI/JNCI-M/JNCI-ER
VP - Technical Training - IPexpert, Inc.
IPexpert Sr. Technical Instructor
[EMAIL PROTECTED]
Telephone: +1.810.326.1444
Fax: +1.810.454.0130
http://www.ipexpert.com
-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Brent Cole
Sent: Thursday, April 17, 2008 1:15 PM
To: [email protected]
Subject: [OSL | CCIE_RS] BGP decision process. Lowest IGP metric or
LowestRID?
BGP stepwise decision process question.
I have two routes to paths in a BGP table and I am trying to determine why
BGP is making its decision the way that it is.
This question relates to task 7.12 in the PG. Here is my "show ip bgp":
Network Next Hop Metric LocPrf Weight Path
* i8.8.8.0/24 150.50.100.5 0 100
0 (65078) ?
*>i 150.50.7.5 0 100
0 (65078) ?
* i150.50.4.0/24 150.50.100.5 0 100
0 (65078) ?
*>i 150.50.7.5 0 100
0 (65078) ?
*> 200.0.0.6/32 0.0.0.0 0 32768 i
For example, looking only at 8.8.8.0/24 for each BGP decision step.
step 1 (next hop reachable)- I can ping both interfaces for NEXT_HOP.
Both are reachable.
R6#ping 150.50.100.5
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 150.50.100.5, timeout is 2
seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max =
64/66/68 ms
R6#ping 150.50.7.5
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 150.50.7.5, timeout is
2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max =
1/3/4 ms
step 2 (IGP synchronization)- synchronization is not enabled so no need for
IGP table sync.
router bgp 65256
no synchronization
bgp log-neighbor-changes
bgp confederation identifier 200
bgp confederation peers 65078
network 200.0.0.6 mask 255.255.255.255
neighbor 150.50.7.5 remote-as 65256
neighbor 150.50.7.5 next-hop-self
neighbor 150.50.100.5 remote-as 65256
no auto-summary
step 3 (Higher Weight)--weights area the same.
see "show ip bgp" above.
step 4 (Higher Local Pref)-- Local_PREF is the same.
see "show ip bgp" above.
step 5 (Locally injected preferred)-- no locally injected routes for this
network
see "router bgp 65256" from step 2 above.
step 6 (AS path legnth)-- as paths area the same.
see "show ip bgp" above.
step 7 (lowest origin type)-- origin types are both incomplete.
see "show ip bgp" above.
step 8 (lowest med value)-- med values area the same.
see "show ip bgp" above.
step 9 (prefer ebgp over others)-- both connections are over iBGP.
see "router bgp 65256" from step 2 above.
step 10 (lowest IGP metric to NEXT_HOP)--both connections area Connected
directly to
NEXT_HOP connection.I have played around with the bandwidth commands on the
interfaces and still seem to get the same path choice. I would guess that
this is where the decision
is different (Ethernet vs. Serial). There is no IGP involved here. Is it
treating the
bandwidth for the interface as the IGP metric?
step 11 (lowest RID)--both router IDs area the same. It is the same neighbor
and thus the same RIDs.
R6#sh ip bgp neigh | include router ID
BGP version 4, remote router ID 200.0.0.5
BGP version 4, remote router ID 200.0.0.5
R6#
The PG states "...if everything is equal, then the path with the lowest
router-id is
preferred. Technically this would always be the ethernet address." As far as
I can
determine, the RID's are the same.
Any clues as to what step BGP is making its decision on in this particular
case?
Thanks in advance,
Brent
