Routing

The routing table lists remote destinations (subnets) with the address of next-hop known to be closer to destination.
On Linux hosts the routing table decides whether the packet should be forwarded or processed locally.
Traditionally the Linux packet forwarding selects the route based on packet's destination IP.
In order to enable forwarding to non-local destination: sysctl -w net.ipv4.ip_forward=1

Routing table

ip route show will display main routing table. Example output:

root@vm1:~# ip r s
default via 192.168.1.1 dev ens3 
10.244.0.0/24 dev cni0 proto kernel scope link src 10.244.0.1 
10.244.1.0/24 via 10.244.1.0 dev flannel.1 onlink 
172.17.0.0/16 dev docker0 proto kernel scope link src 172.17.0.1 linkdown 
192.168.1.0/24 dev ens3 proto kernel scope link src 192.168.1.106 

Dissection: default equals to 0.0.0.0/0
dev <interface> denotes to which interface given route is pinned
via <IP> next hop, may depend on route type src <IP> source address to prefer when using route
proto how the route was installed

proto	info
redirect	installed due to ICMP redirect
kernel	installed by the kernel during autoconfiguration
boot	installed during the bootup sequence
static	installed by administrator
ra	installed by Router Discovery Protocol

scope determines how valid given address is, if not present then global is assumed

scope	info
global	valid everywhere
site	valid only within given site (IPv6)
link	valid only for given device
host	valid only within host, not routable

Routing policy

It is possible to route packet not only based on destination IP, but rather using some additional information, e.g. source address, protocol or packet size.
In order to do that following kernel options must be enabled (y)

> cat /boot/config-$(uname -r) | grep CONFIG_IP_ADVANCED_ROUTER
CONFIG_IP_ADVANCED_ROUTER=y
> cat /boot/config-$(uname -r) | grep CONFIG_XFRM
CONFIG_XFRM=y

Policies can route selectively traffic using different paths.
Policy is picked before routing decision.
Policy can determine which routing table to use (it is possible to have multiple routing tables).

RPDB

Different routing tables are picked based on defined selector.
ip rule list shows which route table will be used for particular set of information

> ip rule list
> #default output
0:      from all lookup local 
32766:  from all lookup main 
32767:  from all lookup default 

> #other host
> ip rule list
0:      from all lookup local 
9:      from all fwmark 0x2 lookup o2 
10:     from all fwmark 0x1 lookup o1 
220:    from all lookup 220 
32766:  from all lookup main 
32767:  from all lookup default 

Dissection: 0: - first column shows entry priority. Lower the number, higher the priority
from all - the selector part: any packet in this case
from all fwmark 0x1 - the selector part: any packet with iptables mark 0x1
lookup tablename - the action part: use tablename routing table

Rules are scanned starting from the lowest number, selector is applied to {src address, dst address, in interface, tos, fwmark}

in order to add routing table to the RPDB (Routing Policy Database): echo "100 o1" >> /etc/iproute2/rt_tables

Display routing table entries: ip r s t tablename

Debugging

ip route get 1.2.3.4 will display which route will be used for particular destination.
ip route get 1.2.3.4 mark 1 will display which route will be used for particular destination with some additional information (mark in this example).

References

1. https://tldp.org/HOWTO/pdf/Adv-Routing-HOWTO.pdf
2. https://silo.tips/download/advanced-routing-scenarios-policy-based-routing-concepts-and-linux-implementatio
3. https://upload.wikimedia.org/wikipedia/commons/3/37/Netfilter-packet-flow.svg
4. http://linux-ip.net/html/tools-ip-rule.html
5. Available IP rules