Dear Maintainer,
*** Reporter, please consider answering these questions, where appropriate ***
This is a known bug, and it leads to total system instability: https://github.com/rfxn/advanced-policy-firewall/issues/48
* What led up to the situation?
I deployed (apt install apf-firewall) apf-firewall to multiple servers and after changing the configuration, I found the firewall did not have my updates.
Update IG__TCP_CPORTS and remove a port. Restart apf (apf -r, or apf -f then apf -s, service restart apf-firewall, systemctl restart apf-firewall) and old rules are are still in IP tables.
* What exactly did you do (or not do) that was effective (or
ineffective)?
I can flush manually as per the github issue: iptables -F, but that broke updates to apf via ansible. Modified the /usr/sbin/apf script to include iptables -F - this broke ansible.
* What was the outcome of this action?
iptables shows that rules were flushed, but the chains still exist. The default policy is reset.
Separate issue: ansible's session is shut down - causing ansible to hang waiting for a response.
* What outcome did you expect instead?
iptables should drop rules, chains, and reset the default policies
*** End of the template - remove these template lines ***