Split brain on keepalived but both hosts get each others messages

Question

I did a little digging and most issues I saw were resolved with a dive into firewalls and multicast fixes but from my pastes below that does not seem to be my issue. I am a little stumped on what it could be at this point.

Debian 9 in vBox Host1: 10.2.0.200 Host2: 10.2.0.201

# tcpdump -i enp0s3 -nn host 10.2.0.201
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on enp0s3, link-type EN10MB (Ethernet), capture size 262144 bytes
10:36:29.992454 IP 10.2.0.201 > 224.0.0.18: VRRPv2, Advertisement, vrid 50, prio 100, authtype none, intvl 1s, length 20
10:36:30.993534 IP 10.2.0.201 > 224.0.0.18: VRRPv2, Advertisement, vrid 50, prio 100, authtype none, intvl 1s, length 20
10:36:31.997348 IP 10.2.0.201 > 224.0.0.18: VRRPv2, Advertisement, vrid 50, prio 100, authtype none, intvl 1s, length 20
10:36:32.998802 IP 10.2.0.201 > 224.0.0.18: VRRPv2, Advertisement, vrid 50, prio 100, authtype none, intvl 1s, length 20

2: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
    link/ether 08:00:27:81:6f:cc brd ff:ff:ff:ff:ff:ff
    inet 10.2.0.200/24 brd 10.2.0.255 scope global enp0s3
       valid_lft forever preferred_lft forever
    inet 10.2.0.50/32 scope global enp0s3
       valid_lft forever preferred_lft forever
    inet6 fe80::a00:27ff:fe81:6fcc/64 scope link
       valid_lft forever preferred_lft forever



# tcpdump -i enp0s3 -nn 'host 10.2.0.200'
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on enp0s3, link-type EN10MB (Ethernet), capture size 262144 bytes
10:35:28.892553 IP 10.2.0.200 > 224.0.0.18: VRRPv2, Advertisement, vrid 50, prio 200, authtype none, intvl 1s, length 20
10:35:29.895166 IP 10.2.0.200 > 224.0.0.18: VRRPv2, Advertisement, vrid 50, prio 200, authtype none, intvl 1s, length 20
10:35:30.895924 IP 10.2.0.200 > 224.0.0.18: VRRPv2, Advertisement, vrid 50, prio 200, authtype none, intvl 1s, length 20
10:35:31.898158 IP 10.2.0.200 > 224.0.0.18: VRRPv2, Advertisement, vrid 50, prio 200, authtype none, intvl 1s, length 20

2: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
    link/ether 08:00:27:2f:ba:89 brd ff:ff:ff:ff:ff:ff
    inet 10.2.0.201/24 brd 10.2.0.255 scope global enp0s3
       valid_lft forever preferred_lft forever
    inet 10.2.0.50/32 scope global enp0s3
       valid_lft forever preferred_lft forever
    inet6 fe80::a00:27ff:fe2f:ba89/64 scope link
       valid_lft forever preferred_lft forever

Ok, so I changed to unicast and now I am getting the error ICMP protocol 112 unreachable. — user206106, Jun 08 '19 at 15:19

score 1 · Answer 1 · answered Jun 08 '19 at 15:48

I am a dummy. First off, even though it was fine in TCPDUMP, making it unicast made me realize there was an iptables issue. Second, when copying iptables rules to fix it, make sure you change your CORRECT interface name so you don't take your frustration out on everyone else because you're a turd.

Split brain on keepalived but both hosts get each others messages

1 Answers1