Floating Rules

Floating Rules are a special type of advanced rule that can perform complicated actions not possible with rules on interface or group tabs. Floating rules can act on multiple interfaces in the inbound, outbound, or both directions. The use of inbound and outbound filtering makes designing the rules more complex and prone to user error, but they can be desirable in specific applications.

Most firewall configurations will never have floating rules, or only have them from the traffic shaper.

Precautions/Caveats

Floating rules can be a lot more powerful than other rules, but also more confusing, and it is easier to make an error that could have unintended consequences in passing or blocking traffic.

Floating rules in the inbound direction, applied to multiple WANs, will not get reply-to added as they would with individual interface rules, so the same problem exists here as existed with interface groups: The traffic will always exit the WAN with the default gateway, and not return properly out the WAN it entered.

Given the relative unfamiliarity of many users with Floating rules, they may not think to look there for rules when maintaining the firewall. As such, they can be a little more difficult for administration since it may not be an obvious place to look for rules.

Be careful when considering the source and destination of packets depending on the inbound and outbound direction. For example, rules in the outbound direction on a WAN would have a local source of the firewall (after NAT) and remote destination.

Potential Uses

The most common use of Floating rules is for ALTQ traffic shaping. Floating tab rules are the only type of rules which can match and queue traffic without explicitly passing the traffic.

Another way to use floating rules is to control traffic leaving from the firewall itself. Floating rules can prevent the firewall from reaching specific IP addresses, ports, and so on.

Other common uses are to ensure that no traffic can exit from other paths into a secure network, no matter what rules exist on other interfaces. By blocking outbound toward a secure network from all but the approved locations, the likelihood of later accidentally allowing traffic in through some other unintended path is reduced. Similarly, they can be used to prevent traffic destined for private networks from leaving a WAN interface, to prevent VPN traffic from leaking.

As mentioned earlier in the interface rules, they can also effectively enact state timeouts, tag/match operations, “no state” rules, and “sloppy state” rules for asymmetric routing.

Processing Order

In the inbound direction, floating rules work essentially the same as interface or group rules except that they are processed first. In the outbound direction, however, things get a little more confusing.

Firewall rules are processed after NAT rules, so rules in the outbound direction on a WAN can never match a local/private IP address source if outbound NAT is active on that interface. By the time it hits the rule, the source address of the packet is now the WAN interface IP address. In most cases this can be worked around by using the match options to tag a packet on the LAN on the way in and then matching that tag on the way out of the firewall.

Floating rules are processed before interface group rules and interface rules, so that must also be taken into consideration.

See also

See Ordering of NAT and Firewall Processing for a more detailed analysis of rule processing and flow through the firewall, including how NAT rules come into play.

Match Action

The match action is unique to floating rules. A rule with the match action will not pass or block a packet, but only match it for purposes of assigning traffic to queues or limiters for traffic shaping. Match rules do not work with Quick enabled.

Quick

Quick controls whether rule processing stops when a rule is matched. The Quick behavior is added to all interface tab rules automatically, but on floating rules it is optional. Without Quick checked, the rule will only take effect if no other rules match the traffic. It reverses the behavior of “first match wins” to be “last match wins”.

Using this mechanism, a default action of sorts can be crafted which will take effect only when no other rules match, similar to the default block rules on WANs.

In most situations, the best practice is to check Quick. There are certain specific scenarios where leaving Quick unchecked is necessary, but they are few and far between. For most scenarios, the only rules they would have without quick selected are match rules traffic shaper rules.

Interface

The Interface selection for floating rules is different than the one for normal interface rules: It is a multi-select box so one, multiple, or all possible interfaces may be selected. Ctrl-click on interfaces to select them one by one, or use other combinations of click/drag or shift-click to select multiple interfaces.

Direction

Floating rules are not limited to the inbound direction like interface rules. They can also act in the outbound direction by selecting out here, or in both directions by selecting any. The in direction is also available.

The out direction is useful for filtering traffic from the firewall itself, for matching other undesirable traffic trying to exit an interface, or for fully configuring “sloppy state” rules, “no state” rules, or alternate state timeouts.

Marking and Matching

Using the Tag and Tagged fields, a connection can be marked by an interface tab rule and then matched in the outbound direction on a floating rule. This is a useful way to act on WAN outbound traffic from one specific internal host that could not otherwise be matched due to NAT masking the source. It can also be used similarly for applying shaping outbound on WAN from traffic specifically tagged on the way into the firewall.

For example, on a LAN rule, use a short string in the Tag field to mark a packet from a source of 10.3.0.56. Then on a floating rule, quick, outbound on WAN, use Tagged with the same string to act on the traffic matched by the LAN rule.