Re: BGP Failure (Syncronisation)

kfarrington · ‎11-18-2003

Hello all,

Can anyone tell me why my iBGP route is not syncronised? Using EIGRP as the IGP, and I can ping my iBGP next hop address?

Many thx indeed,

Ken

LDVD1WR54>sh ip route 128.233.0.0

Routing entry for 128.233.0.0/16

Known via "bgp 500", distance 20, metric 34

Tag 200, type external

Redistributing via eigrp 69

Advertised by eigrp 69

Last update from 192.254.42.254 4d07h ago

Routing Descriptor Blocks:

* 192.254.42.254, from 192.254.42.254, 4d07h ago

Route metric is 34, traffic share count is 1

AS Hops 1

LDVD1WR54>sh ip bgp 128.233.0.0

BGP routing table entry for 128.233.0.0/16, version 1055

Paths: (2 available, best #1, table Default-IP-Routing-Table)

Advertised to non peer-group peers:

141.228.36.129 141.228.61.1 141.228.68.1

200, (received & used)

192.254.42.254 from 192.254.42.254 (31.11.143.2)

Origin IGP, metric 34, localpref 100, valid, external, best

200

192.254.41.254 (metric 30720) from 141.228.36.129 (192.254.41.253)

Origin IGP, metric 23, localpref 100, valid, internal, not synchronized

LDVD1WR54>sh ip route 192.254.41.254

Routing entry for 192.254.41.0/24

Known via "eigrp 69", distance 170, metric 30720, type external

Redistributing via eigrp 69

Last update from 141.228.44.77 on FastEthernet0/0, 3w2d ago

Routing Descriptor Blocks:

* 141.228.45.77, from 141.228.45.77, 3w2d ago, via FastEthernet5/0

Route metric is 30720, traffic share count is 1

Total delay is 200 microseconds, minimum bandwidth is 100000 Kbit

Reliability 179/255, minimum MTU 1500 bytes

Loading 1/255, Hops 1

141.228.44.77, from 141.228.44.77, 3w2d ago, via FastEthernet0/0

Route metric is 30720, traffic share count is 1

Total delay is 200 microseconds, minimum bandwidth is 100000 Kbit

Reliability 179/255, minimum MTU 1500 bytes

Loading 1/255, Hops 1

LDVD1WR54>ping 192.254.41.254

Type escape sequence to abort.

Sending 5, 100-byte ICMP Echos to 192.254.41.254, timeout is 2 seconds:

!!!!!

Success rate is 100 percent (5/5), round-trip min/avg/max = 4/8/12 ms

LDVD1WR54>

ruwhite · ‎11-18-2003

Synchronization isn't related to the reachability of the next hop, but rather the protocol in which the route is learned. In this case, you are learning 128.233.0.0 from an iBGP peer, but you are not learning it in through an interior gateway protocol, as well, so the IGP and BGP tables are not "synchronized."

Synchronization was designed and implemented in a time when the internet routing table was very small, running iBGP full mesh wasn't common, and route reflectors hadn't been invented. The point of synchronization is to prevent a router in the middle of the network which isn't running BGP from receiving traffic that it doesn't know how to forward, if the network is transiting traffic. For instance, suppose you have this:

A----B----C----D

And A and C are exchanging routes over iBGP. Suppose C learns some route, 10.1.1.0/24, through D, an eBGP peer. It readvertises this route to A. Now A receives a packet destined to 10.1.1.1, and sends it on to B, since that's the next hop towards C, where it learned about 10.1.1.0/24. But what is B going to do with the packet? It has no route towards 10.1.1.0/24, since it isn't running BGP at all.

To resolve this, the IGP and BGP tables were "synchronized," by redistributing from BGP into the IGP at C. If the tables were not synchronized, then A wouldn't advertise the route to any external peers, so traffic wouldn't be black holed within the AS. Today, we would either run B in the iBGP mesh, or make B a route reflector.

At any rate, in this case, turn sycnhronization off. If you are transiting traffic, then make certain the routers along the transit path are part of the iBGP mesh. If you aren't then turn synchronization off, and use filters to make certain you aren't advertising anything towards a peer that would cause you to transit traffic.

For more information on synchronization, see:

http://www.cisco.com/en/US/tech/tk365/tk80/technologies_tech_note09186a00800c95bb.shtml#synch

:-)

Russ.W

kfarrington · ‎11-24-2003

Russ,

Please can you checkout the IGP topology table, and you can see that I am learning the route via an IGP but the route appears to be un-syncd. I cannot understand this behaviour?

You can see by the orig router-id that the routes are being learned by different routers, bu the reason the MED is being ignored, is becasue the bgp RIB entry is un-syned.

Please could you help further to explain, if this could be an IOS bug or there is something fundemental I am missing here.

It is causing sub-optimal routing and I need to show my configs to someone (maybe offline) if possible.

Am I going mad here?

Many thx

LDRD1WR04>sh ip eigrp top 121.20.0.0 255.255.0.0

IP-EIGRP topology entry for 121.20.0.0/16

State is Passive, Query origin flag is 1, 1 Successor(s), FD is 2585600

Routing Descriptor Blocks:

0.0.0.0, from Redistributed, Send flag is 0x0

Composite metric is (2585600/0), Route is External

Vector metric:

Minimum bandwidth is 1000 Kbit

Total delay is 1000 microseconds

Reliability is 255/255

Load is 1/255

Minimum MTU is 1500

Hop count is 0

External data:

Originating router is 155.195.44.77 (this system)

AS number of route is 500

External protocol is BGP, external metric is 23

Administrator tag is 200 (0x000000C8)

155.195.44.74 (FastEthernet0/0), from 155.195.44.74, Send flag is 0x0

Composite metric is (2588160/2585600), Route is External

Vector metric:

Minimum bandwidth is 1000 Kbit

Total delay is 1100 microseconds

Reliability is 255/255

Load is 1/255

Minimum MTU is 1500

Hop count is 1

External data:

Originating router is 155.195.44.74

AS number of route is 500

External protocol is BGP, external metric is 34

Administrator tag is 200 (0x000000C8)

155.195.45.74 (FastEthernet5/0), from 155.195.45.74, Send flag is 0x0

Composite metric is (2588160/2585600), Route is External

Vector metric:

Minimum bandwidth is 1000 Kbit

Total delay is 1100 microseconds

Reliability is 255/255

Load is 1/255

Minimum MTU is 1500

Hop count is 1

External data:

Originating router is 155.195.44.74

AS number of route is 500

External protocol is BGP, external metric is 34

Administrator tag is 200 (0x000000C8)

LDRD1WR04>

LDRD1WR04>sh ip bgp 121.20.0.0

BGP routing table entry for 121.20.0.0/16, version 1289

Paths: (2 available, best #2, table Default-IP-Routing-Table)

Advertised to non peer-group peers:

155.195.32.65 155.195.61.1 155.195.68.1

200

192.254.42.254 (metric 30720) from 155.195.32.65 (155.195.32.65)

Origin IGP, metric 34, localpref 100, valid, internal, not synchronized

200, (received & used)

192.254.41.254 from 192.254.41.254 (31.129.255.3)

Origin IGP, metric 23, localpref 100, valid, external, best

LDRD1WR04>

LDVD1WR54>sh ip eigrp top 121.20.0.0 255.255.0.0

IP-EIGRP topology entry for 121.20.0.0/16

State is Passive, Query origin flag is 1, 1 Successor(s), FD is 2585600

Routing Descriptor Blocks:

0.0.0.0, from Redistributed, Send flag is 0x0

Composite metric is (2585600/0), Route is External

Vector metric:

Minimum bandwidth is 1000 Kbit

Total delay is 1000 microseconds

Reliability is 255/255

Load is 1/255

Minimum MTU is 1500

Hop count is 0

External data:

Originating router is 155.195.44.74 (this system)

AS number of route is 500

External protocol is BGP, external metric is 34

Administrator tag is 200 (0x000000C8)

155.195.44.77 (FastEthernet0/0), from 155.195.44.77, Send flag is 0x0

Composite metric is (2588160/2585600), Route is External

Vector metric:

Minimum bandwidth is 1000 Kbit

Total delay is 1100 microseconds

Reliability is 255/255

Load is 1/255

Minimum MTU is 1500

Hop count is 1

External data:

Originating router is 155.195.44.77

AS number of route is 500

External protocol is BGP, external metric is 23

Administrator tag is 200 (0x000000C8)

155.195.45.77 (FastEthernet5/0), from 155.195.45.77, Send flag is 0x0

Composite metric is (2588160/2585600), Route is External

Vector metric:

Minimum bandwidth is 1000 Kbit

Total delay is 1100 microseconds

Reliability is 255/255

Load is 1/255

Minimum MTU is 1500

Hop count is 1

External data:

Originating router is 155.195.44.77

AS number of route is 500

External protocol is BGP, external metric is 23

Administrator tag is 200 (0x000000C8)

LDVD1WR54>sh ip bgp 121.20.0.0

BGP routing table entry for 121.20.0.0/16, version 1238

Paths: (2 available, best #1, table Default-IP-Routing-Table)

Advertised to non peer-group peers:

155.195.36.129 155.195.61.1 155.195.68.1

200, (received & used)

192.254.42.254 from 192.254.42.254 (31.11.143.2)

Origin IGP, metric 34, localpref 100, valid, external, best

200

192.254.41.254 (metric 30720) from 155.195.36.129 (192.254.41.253)

Origin IGP, metric 23, localpref 100, valid, internal, not synchronized

LDVD1WR54>

ruwhite · ‎11-24-2003

Is the route that's not synch'd 121.20.0.0, or one of the other ones you posted before? Could you post a show ip route x.x.x.x, show ip eigrp topo x.x.x.x, and a show ip bgp x.x.x.x for one of the routes that isn't synch'd?

:-)

Russ.W

kfarrington · ‎11-24-2003

Hi Russ,

It is the one I just posted, but have many routes like this.

Here is the output, show ip eigrp top, show ip bgp and show ip route for both routers. Both routers connect to AS200 ebgp and ibgp of AS500

The way I understand this should work, is for both router 1 and router 2 to accept the external route and then compare MED. Both router 1 and router 2 receive this route via ibgp also (lets say before split horizon), The prefix in Router 1 has the best MED and router 2 will have the 2 entries for the prefix, one external, and one internal and should chose lowest MED before the iBGP/eBGP descision so chooses the iBGP path. Router1 should advertised this into the IGP via redistribution and router 2 should not.. Also, Some form of "Cisco BGP split horizon" is employed so rtr2 does not advertise the higher med route it learned via eBGP over the ibgp link to rtr1. (i beleieve this is Cisco spec...)

Please tell me if I am way off this one :)

Sorry about the one sentance but was remembering the process while writing this.

Many thx

-----------------Router 1----------------------------------

LDRD1WR04>sh ip eigrp top 121.20.0.0 255.255.0.0

IP-EIGRP topology entry for 121.20.0.0/16

State is Passive, Query origin flag is 1, 1 Successor(s), FD is 2585600

Routing Descriptor Blocks:

0.0.0.0, from Redistributed, Send flag is 0x0

Composite metric is (2585600/0), Route is External

Vector metric:

Minimum bandwidth is 1000 Kbit

Total delay is 1000 microseconds

Reliability is 255/255

Load is 1/255

Minimum MTU is 1500

Hop count is 0

External data:

Originating router is 155.195.44.77 (this system)

AS number of route is 500

External protocol is BGP, external metric is 23

Administrator tag is 200 (0x000000C8)

155.195.44.74 (FastEthernet0/0), from 155.195.44.74, Send flag is 0x0

Composite metric is (2588160/2585600), Route is External

Vector metric:

Minimum bandwidth is 1000 Kbit

Total delay is 1100 microseconds

Reliability is 255/255

Load is 1/255

Minimum MTU is 1500

Hop count is 1

External data:

Originating router is 155.195.44.74

AS number of route is 500

External protocol is BGP, external metric is 34

Administrator tag is 200 (0x000000C8)

155.195.45.74 (FastEthernet5/0), from 155.195.45.74, Send flag is 0x0

Composite metric is (2588160/2585600), Route is External

Vector metric:

Minimum bandwidth is 1000 Kbit

Total delay is 1100 microseconds

Reliability is 255/255

Load is 1/255

Minimum MTU is 1500

Hop count is 1

External data:

Originating router is 155.195.44.74

AS number of route is 500

External protocol is BGP, external metric is 34

Administrator tag is 200 (0x000000C8)

LDRD1WR04>

LDRD1WR04>sh ip bgp 121.20.0.0

BGP routing table entry for 121.20.0.0/16, version 1289

Paths: (2 available, best #2, table Default-IP-Routing-Table)

Advertised to non peer-group peers:

155.195.32.65 155.195.61.1 155.195.68.1

200

192.254.42.254 (metric 30720) from 155.195.32.65 (155.195.32.65)

Origin IGP, metric 34, localpref 100, valid, internal, not synchronized

200, (received & used)

192.254.41.254 from 192.254.41.254 (31.129.255.3)

Origin IGP, metric 23, localpref 100, valid, external, best

LDRD1WR04>

LDRD1WR04>sh ip route 121.20.0.0

Routing entry for 121.20.0.0/16

Known via "bgp 500", distance 20, metric 23

Tag 200, type external

Redistributing via eigrp 69

Advertised by eigrp 69

Last update from 192.254.41.254 4d18h ago

Routing Descriptor Blocks:

* 192.254.41.254, from 192.254.41.254, 4d18h ago

Route metric is 23, traffic share count is 1

AS Hops 1

LDRD1WR04>

-----------------Router 2----------------------------------

LDVD1WR54>sh ip eigrp top 121.20.0.0 255.255.0.0

IP-EIGRP topology entry for 121.20.0.0/16

State is Passive, Query origin flag is 1, 1 Successor(s), FD is 2585600

Routing Descriptor Blocks:

0.0.0.0, from Redistributed, Send flag is 0x0

Composite metric is (2585600/0), Route is External

Vector metric:

Minimum bandwidth is 1000 Kbit

Total delay is 1000 microseconds

Reliability is 255/255

Load is 1/255

Minimum MTU is 1500

Hop count is 0

External data:

Originating router is 155.195.44.74 (this system)

AS number of route is 500

External protocol is BGP, external metric is 34

Administrator tag is 200 (0x000000C8)

155.195.44.77 (FastEthernet0/0), from 155.195.44.77, Send flag is 0x0

Composite metric is (2588160/2585600), Route is External

Vector metric:

Minimum bandwidth is 1000 Kbit

Total delay is 1100 microseconds

Reliability is 255/255

Load is 1/255

Minimum MTU is 1500

Hop count is 1

External data:

Originating router is 155.195.44.77

AS number of route is 500

External protocol is BGP, external metric is 23

Administrator tag is 200 (0x000000C8)

155.195.45.77 (FastEthernet5/0), from 155.195.45.77, Send flag is 0x0

Composite metric is (2588160/2585600), Route is External

Vector metric:

Minimum bandwidth is 1000 Kbit

Total delay is 1100 microseconds

Reliability is 255/255

Load is 1/255

Minimum MTU is 1500

Hop count is 1

External data:

Originating router is 155.195.44.77

AS number of route is 500

External protocol is BGP, external metric is 23

Administrator tag is 200 (0x000000C8)

LDVD1WR54>sh ip bgp 121.20.0.0

BGP routing table entry for 121.20.0.0/16, version 1238

Paths: (2 available, best #1, table Default-IP-Routing-Table)

Advertised to non peer-group peers:

155.195.36.129 155.195.61.1 155.195.68.1

200, (received & used)

192.254.42.254 from 192.254.42.254 (31.11.143.2)

Origin IGP, metric 34, localpref 100, valid, external, best

200

192.254.41.254 (metric 30720) from 155.195.36.129 (192.254.41.253)

Origin IGP, metric 23, localpref 100, valid, internal, not synchronized

LDVD1WR54>

LDVD1WR54>sh ip route 121.20.0.0

Routing entry for 121.20.0.0/16

Known via "bgp 500", distance 20, metric 34

Tag 200, type external

Redistributing via eigrp 69

Advertised by eigrp 69

Last update from 192.254.42.254 4d18h ago

Routing Descriptor Blocks:

* 192.254.42.254, from 192.254.42.254, 4d18h ago

Route metric is 34, traffic share count is 1

AS Hops 1

LDVD1WR54>

ruwhite · ‎11-24-2003

BGP looks in the routing table to determine sycnhronization, not the EIGRP table, so... If the EIGRP route was in the routing table, then the iBGP route should be synchronized, and it would be advertised to external peers. Since the eBGP learned route is in the routing table, then EIGRP route isn't, and the iBGP route shows not synchronized.

:-)

Russ.W

kfarrington · ‎11-24-2003

Hey Russ,

As the route shows up in the topology table,AD of 170, and the route in the routing table is eBGP - AD of 20, eBGP is always gonna win for insertion into the forwarding table. But......

if the route was learned by EIGRP before the eBGP route, the route should have an AD of iBGP (200) and thus EIGRP should win.

How does this work, and is it "purley order of operation" that determines which route should be in the routing table.

Do you get my drift mate?

Regards, as always,

Ken

ruwhite · ‎11-24-2003

It's an order of operation/internal BGP issue--the eBGP route will always win, as long as it's present. The EIGRP route should win if only the iBGP route is present.

:-)

Russ.W

kfarrington · ‎11-25-2003

Russ,

Thx for the help mate.

It is becuase our MED is note being compared that the BGP tables are screwed up. As we receive the saem route from two eBGP routers, and peer iBGP, the local BGP descision process should use the route with the lowest med (say thats via iBGP, and not use the eBGP route, but this is not occuring.

I will pick this up with our Cisco Advanced Services guys unless you have a beeter avenue of approach.

Many thx dude,

Kem

ruwhite · ‎11-25-2003

Are you using deterministic med? It might work better, since there are some issues with path ordering (the order in which routes are received) when comparing routes, causing a route with worse med to be preferred over a route with better med.

http://www.cisco.com/en/US/partner/tech/tk365/tk80/technologies_tech_note09186a0080094925.shtml

:-)

Russ.W