Open Infrastructure

• router2.place6 takes care of 100% of outgoing IPv6 traffic
• Phasing in IPv4 routing by preferring parts of the Internet step by step (see below)
  • We split the IPv4 Internet into 8 parts and change 1/8 of the traffic each time to go onto router2
  • Meanwhile running perf to find out the bottleneck
  • Might be related to https://www.linuxquestions.org/questions/linux-server-73/high-cpu-load-by-native_queued_spin_lock_slowpath-4175615354/

[14:08] line:~% sipcalc -s /3 0.0.0.0/0  
-[ipv4 : 0.0.0.0/0] - 0

[Split network]
Network            - 0.0.0.0         - 31.255.255.255
Network            - 32.0.0.0        - 63.255.255.255
Network            - 64.0.0.0        - 95.255.255.255
Network            - 96.0.0.0        - 127.255.255.255
Network            - 128.0.0.0       - 159.255.255.255
Network            - 160.0.0.0       - 191.255.255.255
Network            - 192.0.0.0       - 223.255.255.255
Network            - 224.0.0.0       - 255.255.255.255

• 2019-09-18, ca. 1530 KST: enabled 0.0.0.0/3 to go through router2.place6 with full routing table
• 2019-09-18, ca. 1600 KST: enabled 32.0.0.0/3 to go through router2.place6 with full routing table

Configuring network on router2.place5:

# Sunrise
auto bond0.102
iface bond0.102 inet6 static
  address 2001:1700:3500::12
  netmask 64
  post-up /sbin/ip link set $IFACE mtu 9000
  vlan-raw-device bond0

iface bond0.102 inet static
  address 195.141.230.103
  netmask 255.255.255.254
  vlan-raw-device bond0

[18:54:33] router2.place5:/etc/network/interfaces.d# ifup bond0.102                                                           
[18:54:42] router2.place5:/etc/network/interfaces.d# 

Switch config

Et41                      connected    1        a-full a-1G   1000BASE-SX   

switch1-place5(config)#interface ethernet 41
switch1-place5(config-if-Et41)#switchport trunk allowed vlan 101,102
switch1-place5(config-if-Et41)#switchport mode trunk 
switch1-place5(config-if-Et41)#write 
Copy completed successfully.
switch1-place5(config-if-Et41)#

switch1-place5(config-if-Et41)#vlan 101
switch1-place5(config-vlan-101)#name sunrise_router1
switch1-place5(config-vlan-101)#vlan 102
switch1-place5(config-vlan-102)#name sunrise_router2
switch1-place5(config-vlan-102)#write 
Copy completed successfully.

Receiving packets from peer:

64 bytes from 2001:1700:3500::11: icmp_seq=296 ttl=64 time=0.977 ms
64 bytes from 2001:1700:3500::11: icmp_seq=297 ttl=64 time=1.01 ms
64 bytes from 2001:1700:3500::11: icmp_seq=298 ttl=64 time=0.988 ms
64 bytes from 2001:1700:3500::11: icmp_seq=299 ttl=64 time=0.997 ms
64 bytes from 2001:1700:3500::11: icmp_seq=300 ttl=64 time=1.00 ms

64 bytes from 195.141.230.102: icmp_seq=474 ttl=255 time=0.937 ms
64 bytes from 195.141.230.102: icmp_seq=475 ttl=255 time=0.943 ms

Saving config, updating with new bgp type

[19:04:41] router2.place5:/etc/bird# rsync -av /etc/bird/ ~/bird20190921
sending incremental file list
created directory /root/bird20190921
./
bird.conf
bird.conf.ucf-dist
bird6.conf
bird6.conf.ucf-dist
envvars
general.conf
init7_v4.conf
init7_v6.conf
router1_place5_ungleich_ch_v4.conf
router1_place5_ungleich_ch_v6.conf
router1_place6_ungleich_ch_v4.conf
router1_place6_ungleich_ch_v6.conf
router2_place6_ungleich_ch_v4.conf
router2_place6_ungleich_ch_v6.conf
router_id.conf
switch1_place10_ungleich_ch_v6.conf
switch1_place5_ungleich_ch_v6.conf
switch1_place6_ungleich_ch_v6.conf
switch1_place8_ungleich_ch_v6.conf
switch2_place5_ungleich_ch_v6.conf
switch2_place6_ungleich_ch_v6.conf
switch2_place8_ungleich_ch_v6.conf
switch3_place6_ungleich_ch_v6.conf
switch4_place6_ungleich_ch_v6.conf
switch5_place6_ungleich_ch_v6.conf
switch6_place6_ungleich_ch_v6.conf
ungleich-routes-place5_v4.conf
ungleich-routes-place5_v6.conf

sent 21,681 bytes  received 592 bytes  44,546.00 bytes/sec
total size is 19,302  speedup is 0.87
[19:04:48] router2.place5:/etc/bird#