kURL: Diagnosing ip_forward being disabled

Troubleshooting
1

Hello,

I have a VM that has a couple of errors while doing the upgrade:

Error from server (BadRequest): container "kotsadm" in pod "kotsadm-0" is waiting to start: PodInitializing

or

Error from server (BadRequest): container "restore-db" in pod "kotsadm-57bd644c6f-f2ngx" is waiting to start: PodInitializing

Also, in coredns logs I see the following errors

failed to connect to postgres: failed to connect to postgres: failed to connect to `host=kotsadm-postgres user=kotsadm database=kotsadm`: hostname resolving error (lookup kotsadm-postgres on 10.100.0.10:53: dial udp 10.100.0.10:53: i/o timeout)
Error: failed to connect to postgres: failed to connect to postgres: failed to connect to `host=kotsadm-postgres user=kotsadm database=kotsadm`: hostname resolving error (lookup kotsadm-postgres on 10.100.0.10:53: dial udp 10.100.0.10:53: i/o timeout)

What could be causing these issues?

2

Generally, but not always, this happens when IP forward is disabled which is required for Docker to operate properly.

In order to diagnose this, run a few of the following commands:

  • systemctl status firewalld

    • Check to see if firewalld is disabled.

  • iptables -L -v

    • Double check to see no firewall rules are blocking.

  • docker info

    • At the bottom of the output see if there are any WARNINGs related to IPv4 forward being disabled.

  • sysctl -a

    • verify this is set to net.ipv4.ip_forward = 1
    • If its set to 0, set it to 1 using sysctl net.ipv4.ip_forward=1
    • Check to make sure /etc/sysctl.conf is also set to net.ipv4.ip_forward=1.