Please enable JavaScript.
Coggle requires JavaScript to display documents.
Route Selection Route Manipulation (Redistribution between mergers with…
Route Selection
Route Manipulation
Route Selection
HINT: Packet Switching == layer 3
Frame Switching == layer 2
PBR (Policy Based Routing)
overrule Routing table decisions
2- create named Route-map
match AC or Prefix list
set next-hop[
- set without default:
try to use PBR , if it is not working , use normal routing
- set with default:
use the normal routing , if it isn't working, then use PBR
]
3- apply Route-map:
ip local policy route-map
for originating traffic from router in global config mode
-
ip policy Route-map
for traffic flow through the router under interface mode
1- create named AC or Prefix-list
IP SLA (IP Service Level Agreement)
monitor the network and change Routing decisions, use
object Tracking
to check reachability of IP Address(by pinging) or certain service by connecting to it using TCP.
verify-availability parameter to use PBR with IP SLA
conf of non-cisco device:
-tcp-connect 10.0.0.1 23 control disable
available operations types:
ICMP(echo, jitter)
UDP(echo, jitter)
HTTP
TCP connections
RTP(voice)
DNS
DHCP
FTP
to configure PBR to use object-oriented Tracking:
set ip next-hop
verify-availability
10.1.14.4 1 track 2
3- CEF (Cisco Express Forwarding)
allow really fast Forwarding, default for routers
enabled globally be
ip cef
command
unter interface by
ip route-cache cef
command
Control Plane(BRAIN)
for
exchanging
routing information using routing protocols and
forming
Routing table and ARP Table, routing descisions, can't forward packets because of
Recursive Routing
Exceptions forwarded by control plane:
packets destined to one of the IP Addresses of the router
Routing protocol Traffic like OSPF, EIGRP
IP Packets have some options in IP header
IP Packets with expired TTL
Data Plane(MUSCLES)
for Forwarding IP Packets using
ASICS
, that stores Routing tables, ARP tables and Access Lists, builds
Hardware Tables
with them
CAM:
Content Addressable Memory
stores Layer 2 Infos, like MACs, MAC Interfaces, VLANs
exact Matching informatons
Adjacency Table
sh adjacency summary
Hardware table based on ARP Table
Header Packet Information:
Source MAC
Destination MAC
IP Packet TTL
IP Packet Checksum
Ethernet Frame Checksum
TCAM:
Ternary Content Addressable Memory
stores Higher Layer Informations, like AC, QoS, and Routing Tables
Not exact Matching Information
FIB(Forwarding Information Base)
sh ip cef
Hardware table based on Routing table
Valid Cache Adjacency:
for successful packets forwarding
Null Adjacency:
packets to Null0 intrface
Drop Adjacency:
packets have encapsulation error, or route can't be resolved
Discard Adjacency:
packets discarded because of AC or Policy map
Punt Adjacency:
packets sent to control plane for processing
Glean Adjacency:
Router has one subnet-prefix of the all directly connected hosts
2- Fat Switching
more efficient, cashes forwarding information to use it again.
disabled by
no ip route cache
1- Process Switching
very slow, because forwarding done with CPU(software)
Router remove ethernet frame and look up destination ip address then forward with the rewritten MAC Address
Management plane
to manage the router through SSH, VTTY
VRF (Virtual Routing and Forwarding)
allow dividing Routing table into multiple virtual Routers, each Router with its own Routing table.
it is usually used with MPLS
Layer 3 Interface can't belong to more than one VRF
VRF lite
VRF Deployment without MPLS, used with catalyst switsches to support multiple VPN Instances in customer Edge
created with
ip vrf command + name
apply the vrf under the interface with
ip vrf forwarding + VRF name
command
it becomes complex with more router using multiple VRFs
requires subinterfaces for vrf on each device
to keep voice and video separatedeperated from data
the OSPF process ID needs to be unique for each VRF.
EVN(Easy Virtual Networking)
removes the need of per-vrf suninterfaces with VRF-Lite
uses
Virtual Network Trunk(Vnet Trunk)
Vnet Trunk carries all Vrf Traffic
Routers will be connected to Switches using 802.1Q trunk for VLANS
it makes it possible to share service from one network with another
it is useful for shared services like DNS and DHCP
Sharing done with
Route Replication
Service
Route Replication allows each virtual Network to access the RIB of each VRF
Rout Manipulation
Distribute list with prefix-list:
for specific prefixes
Distribute list route-map
identify subset of routes to identify or change based on the prefix or other factors
make filtering choices about which route to be redistributed and which not
set the metric to different values based on the factors matched by the route-map
set the type of external routes in OSPF
Match command
refer to another tools like ACL or Prefix-list to match anything matchable by these tools
can match other parameters directly
1- match interface
2- match ip address [ACl Name/ Number] or prefix-list [prefix-listnale]
3- match ip next-hop[ACl Name/ Number]
4- match ip route-source [ACl Name/ Number]
5- match metric
6- match route-type[Internal/ external]
7-match tag-value(which be taged from other router)
Distribute list with AC
Redistribution
between mergers with different IGPs
between mergers with same IGP(like EIGRP with different ASN)
Between Routing protocols
static Routes into routing protocols
directly connected routes on interfaces on which protocol enabled
between IGP and BGP in MPLS
it merges the routing tables, not the topology tables
*
You cannot redistribute
what you don’t have in your routing table
One-way Redistribution:
Redistribution occurs in only one direction
static route has to be configured from other direction for full connectivity
One-way multipoint redistribution:
occurs in one direction from two routers/ backup/ not full routing protocols exchange/ sub-optimal routing/Routing Domain Loops
Tow-way Redistribution:
between two protocols from one single router(One-point redistribution)/ no loops/ no backups
Tow-way multipoint redistribution
occurs in two direction from two routers/ backup/ full routing protocols connectivity/ probably loops/suboptimal routing
Requirements:
one physical link between the two domains
working routing protocols in each domain
apply redistribute command
Changing Metric:
-
default metric
command
change parameters after
redistribute command
using
route-map
Default seed metric:
RIP
: infinity : :
EIGRP
: infinity
OSPF
: default 20/ BGP 1/ same metric form other OSPF Process
BGP
: set to IGP metric
OSPF by default prefers internal Metrics than External and E1 over E2