Difference between revisions of "Route optimization"

From Teknologisk videncenter
Jump to: navigation, search
m (New page: = reditribution cavets = == Routing loops == === redistribute RIP into OSPF example === Consider the network below and follow the red update path. #Router A updates router B with RIP. Rou...)
 
m (→‎prefix lists)
 
(8 intermediate revisions by the same user not shown)
Line 1: Line 1:
−
= reditribution cavets =
+
= redistribution cavets =
 
== Routing loops ==
 
== Routing loops ==
 
=== redistribute RIP into OSPF example ===
 
=== redistribute RIP into OSPF example ===
 
Consider the network below and follow the red update path.  
 
Consider the network below and follow the red update path.  
 
#Router A updates router B with RIP. Router B now learne the network 80.0.0.0/16.  
 
#Router A updates router B with RIP. Router B now learne the network 80.0.0.0/16.  
−
**Please note that the administrative distance for RIP is 120.
+
##Please note that the administrative distance for RIP is 120.
 
#Router B now redistributes 80.0.0.0/16 into OSPF.
 
#Router B now redistributes 80.0.0.0/16 into OSPF.
−
**Note that OSPF has an administrative distance of 110.
+
##Note that OSPF has an administrative distance of 110.
 
#Router B updates router C as an '''external''' OSPF route administrative distance 110
 
#Router B updates router C as an '''external''' OSPF route administrative distance 110
 
#Router C updates router D as an '''external''' OSPF route administrative distance 110
 
#Router C updates router D as an '''external''' OSPF route administrative distance 110
Line 21: Line 21:
 
Consider the network below and follow the red update path.  
 
Consider the network below and follow the red update path.  
 
#Router A updates router B with RIP. Router B now learne the network 80.0.0.0/16.  
 
#Router A updates router B with RIP. Router B now learne the network 80.0.0.0/16.  
−
**Please note that the administrative distance for RIP is 120.
+
##Please note that the administrative distance for RIP is 120.
 
#Router B now redistributes 80.0.0.0/16 into EIGRP.
 
#Router B now redistributes 80.0.0.0/16 into EIGRP.
−
**EIGRP has an '''internal''' route administrative distance of 90. and
+
##EIGRP has an '''internal''' route administrative distance of 90. and
−
**EIGRP has an '''external''' route administrative distance of 170.  
+
##EIGRP has an '''external''' route administrative distance of 170.  
 
#Router B updates router C as an '''external''' EIGRP route administrative distance 170
 
#Router B updates router C as an '''external''' EIGRP route administrative distance 170
 
#Router C updates router D as an '''external''' EIGRP route administrative distance 170
 
#Router C updates router D as an '''external''' EIGRP route administrative distance 170
Line 33: Line 33:
 
#Router E sends traffic to 80.0.0.0/16 to router A because RIP(120) administrative distance is better than EIGRP external(170).  
 
#Router E sends traffic to 80.0.0.0/16 to router A because RIP(120) administrative distance is better than EIGRP external(170).  
 
[[image:RIP_redistribute_into_EIGRP.png|500px|none|thumb|RIP redistribution into EIGRP example]]
 
[[image:RIP_redistribute_into_EIGRP.png|500px|none|thumb|RIP redistribution into EIGRP example]]
 +
 
= Distribute lists =
 
= Distribute lists =
 
You can filter updates with distribute lists. The exampe below shows filtering of incoming updates. You can also filter outgoing updates.
 
You can filter updates with distribute lists. The exampe below shows filtering of incoming updates. You can also filter outgoing updates.
 
[[image:distribute_list.png|500px|none|thumb|distribute list example]]
 
[[image:distribute_list.png|500px|none|thumb|distribute list example]]
−
<pre>
+
<source lang=cli>
 
hostname B
 
hostname B
 
!
 
!
Line 45: Line 46:
 
access-list 23 deny 90.0.0.0 0.0.255.255
 
access-list 23 deny 90.0.0.0 0.0.255.255
 
access-list 23 permit any
 
access-list 23 permit any
−
</pre>
+
</source>
 
Router B Configuration of distribute list
 
Router B Configuration of distribute list
 
<br/><br/>
 
<br/><br/>
−
<pre>
+
<source lang=cli>
 
hostname E
 
hostname E
 
!
 
!
Line 57: Line 58:
 
access-list 3 deny 80.0.0.0 0.0.255.255
 
access-list 3 deny 80.0.0.0 0.0.255.255
 
access-list 3 permit any
 
access-list 3 permit any
−
</pre>
+
</source>
 
Router E Configuration of distribute list
 
Router E Configuration of distribute list
 
<br/><br/>
 
<br/><br/>
 +
=prefix lists=
 +
Routes in routetable
 +
{|border=1 ;style="margin: 0 auto; text-align: center;cellpadding="5" cellspacing="0"
 +
|+ Address groups
 +
|- bgcolor=lightgrey
 +
! Nr. !!Network
 +
|-
 +
| 1 || 10.0.0.0/8
 +
|-
 +
| 2 || 10.128.0.0/9
 +
|-
 +
| 3 || 10.1.1.0/24
 +
|-
 +
| 4 || 10.1.2.0/24
 +
|-
 +
| 5 || 10.128.10.4/30
 +
|-
 +
| 6 || 10.128.10.8/30
 +
|-
 +
|}
 +
 +
{|border=1 ;style="margin: 0 auto; text-align: center;cellpadding="5" cellspacing="0"
 +
|+ Address groups
 +
|- bgcolor=lightgrey
 +
! Prefix list !!Matches in previos table
 +
|-
 +
| 10.0.0.0/8 || 1
 +
|-
 +
| 10.128.0.0/9 || 2
 +
|-
 +
| 10.0.0.0/8 ge 9|| 2,3,4,5,6
 +
|-
 +
| 10.0.0.0/8 ge 24 le 24 || 3,4
 +
|-
 +
| 10.0.0.0/8 le 24 || 1,2,3,4
 +
|-
 +
| 0.0.0.0/0 || None
 +
|-
 +
| 0.0.0.0/0 le 32 || All
 +
|}
 +
{{:Prefix-list Cisco IOS}}
 +
 
= PBR: Policy Based Routing =
 
= PBR: Policy Based Routing =
 
PBR or Policy Based Routing gives more control and more options than distribute lists - see above.
 
PBR or Policy Based Routing gives more control and more options than distribute lists - see above.
Line 65: Line 108:
 
To select which packets should go where you can use the '''match''' command and match the based on for example source or destination addresses.
 
To select which packets should go where you can use the '''match''' command and match the based on for example source or destination addresses.
  
−
In the picture below the traffic from 192.168.1.0724 on E0 interface must route to ISP1 and
+
In the picture below the traffic from 192.168.1.0/24 on E0 interface must route to ISP1 and
 
traffic from 172.16.1.0/24 on E1 must route to ISP1.
 
traffic from 172.16.1.0/24 on E1 must route to ISP1.
−
<pre>
+
<source lang=cli>
 
hostname RTA
 
hostname RTA
 
!
 
!
Line 75: Line 118:
 
route-map ISP1 permit 10
 
route-map ISP1 permit 10
 
  match ip address 37
 
  match ip address 37
−
  set interface s0/0
+
  set interface s0
 
!
 
!
 
route-map ISP2 permit 20
 
route-map ISP2 permit 20
 
  match ip address 48
 
  match ip address 48
−
  set interface s0/1
+
  set interface s1
 
!
 
!
 
interface e0
 
interface e0
Line 86: Line 129:
 
interface e1
 
interface e1
 
  ip policy route-map ISP2
 
  ip policy route-map ISP2
−
</pre>
+
</source>
 
   
 
   
 
[[Image:PBR1.png|500px|none|thumb|Policy Based Routing example]]
 
[[Image:PBR1.png|500px|none|thumb|Policy Based Routing example]]
  
 +
= Seed metric =
 +
{|
 +
|[[Image:Seed metric1.gif|600px|thumb|Default Seed Metric]]
 +
|-
 +
|[[Image:Seed metric2.gif|600px|thumb|Default Seed Metric example (RIP=infinity default)]]
 +
|-
 +
|[[Image:Seed metric3.gif|600px|thumb|Administrative distance]]
 +
|-
 +
|}
  
−
[[Category:Network]][[Category:Cisco]]
+
{{source cli}}
 +
[[Category:Cisco]][[Category:CCNP1]][[Category:CCNP]][[Category:IOS]][[Category:Network]]

Latest revision as of 13:29, 16 August 2022

redistribution cavets

Routing loops

redistribute RIP into OSPF example

Consider the network below and follow the red update path.

  1. Router A updates router B with RIP. Router B now learne the network 80.0.0.0/16.
    1. Please note that the administrative distance for RIP is 120.
  2. Router B now redistributes 80.0.0.0/16 into OSPF.
    1. Note that OSPF has an administrative distance of 110.
  3. Router B updates router C as an external OSPF route administrative distance 110
  4. Router C updates router D as an external OSPF route administrative distance 110
  5. Router D updates router E as an external OSPF route administrative distance 110
  6. Router E now has two router to 80.0.0.0/16
    1. Via RIP directly to Router A (Preferred route)
    2. Via OSPF to router D - Path D->C->B->A.
  7. Router E sends traffic to 80.0.0.0/16 to router D because OSPF(110) administrative distance is better than RIP's(120).
RIP redistribution into OSPF example

Avoiding Routing loop

redistribute RIP into EIGRP example

With EIGRP we avoid the routing loop seen from OSPF above. EIGRP has two different administrative distances. One for internal routes 90 and one for externally learned routes 170.
Consider the network below and follow the red update path.

  1. Router A updates router B with RIP. Router B now learne the network 80.0.0.0/16.
    1. Please note that the administrative distance for RIP is 120.
  2. Router B now redistributes 80.0.0.0/16 into EIGRP.
    1. EIGRP has an internal route administrative distance of 90. and
    2. EIGRP has an external route administrative distance of 170.
  3. Router B updates router C as an external EIGRP route administrative distance 170
  4. Router C updates router D as an external EIGRP route administrative distance 170
  5. Router D updates router E as an external EIGRP route administrative distance 170
  6. Router E now has two router to 80.0.0.0/16
    1. Via RIP directly to Router A (Preferred route)
    2. Via EIGRP to router D - Path D->C->B->A.
  7. Router E sends traffic to 80.0.0.0/16 to router A because RIP(120) administrative distance is better than EIGRP external(170).
RIP redistribution into EIGRP example

Distribute lists

You can filter updates with distribute lists. The exampe below shows filtering of incoming updates. You can also filter outgoing updates.

distribute list example
hostname B
!
router rip
 version 2
 distribute-list 23 in
!
access-list 23 deny 90.0.0.0 0.0.255.255
access-list 23 permit any

Router B Configuration of distribute list 

hostname E
!
router rip
 version 2
 distribute-list 3 in
!
access-list 3 deny 80.0.0.0 0.0.255.255
access-list 3 permit any

Router E Configuration of distribute list

prefix lists

Routes in routetable

Address groups
Nr. Network
1 10.0.0.0/8
2 10.128.0.0/9
3 10.1.1.0/24
4 10.1.2.0/24
5 10.128.10.4/30
6 10.128.10.8/30
Address groups
Prefix list Matches in previos table
10.0.0.0/8 1
10.128.0.0/9 2
10.0.0.0/8 ge 9 2,3,4,5,6
10.0.0.0/8 ge 24 le 24 3,4
10.0.0.0/8 le 24 1,2,3,4
0.0.0.0/0 None
0.0.0.0/0 le 32 All

Example

Prefix-lists are numbered. In this example "secret net" are not announced.

  • Note: As with access-lists there are an implicit deny in the end.
ip prefix-list SECRET-NET seq 10 deny 172.16.0.0/16
ip prefix-list SECRET-NET seq 20 deny 192.168.22.0/24
ip prefix-list SECRET-NET seq 30 permit 0.0.0.0/0 le 32 !Permit all other nets
!
router bgp 100
  neighbor 10.1.2.3 remote-as 200
  neighbor 10.1.2.3 prefix-list SECRET-NET out

PBR: Policy Based Routing

PBR or Policy Based Routing gives more control and more options than distribute lists - see above. PBR introduces the set command which set the interface or next-hop address to which the packet should be sent. To select which packets should go where you can use the match command and match the based on for example source or destination addresses.

In the picture below the traffic from 192.168.1.0/24 on E0 interface must route to ISP1 and traffic from 172.16.1.0/24 on E1 must route to ISP1. 

hostname RTA
!
access-list 37 permit 192.168.1.0 0.0.0.255
access-list 48 permit 172.16.1.0 0.0.0.255
!
route-map ISP1 permit 10
 match ip address 37
 set interface s0
!
route-map ISP2 permit 20
 match ip address 48
 set interface s1
!
interface e0
 ip policy route-map ISP1
!
interface e1
 ip policy route-map ISP2
Policy Based Routing example

Seed metric

Default Seed Metric
Default Seed Metric example (RIP=infinity default)
Administrative distance
Retrieved from "http://mars.merhot.dk/w/index.php?title=Route_optimization&oldid=40669"