By default when adding the interception rules, we were killing all
existing connections, to force them go to the netfilter queue.
However in some environments this is not acceptable, so now it's configurable.
Besides, we were doing this only for nftables, so now it also works for
iptables.
This option will only take effect if FlushConnsOnStart is true, and the
user has (re)configured the ProcMonitorMethod option (typically on
start, or when changing it manually).
cherry picked from commits:
8e9c1d21788935bfe6b0
related commits:
661e3da48f1e6d2d464b
When we start to intercept connections, we flush out the conntrack
table, to force already established connections reconnect again so we
can intercept them, and let the user choose if allow or deny them.
Since we no longer use conntrack states to intercept TCP connections, we
now close existing connections, leaving to the applications reestablish
them again.
Local connections are excluded, because it may cause problems with some
local servers.
Both options interfere with the established connections, so you may
experience ocasional network interruptions when enabling the
interception for the first time.
Discussion: #995
Using "ct state NEW" to intercept packets causes some undesired effects:
We intercept packets that not only have the SYN flag set, like ACK,
ACK+PSH or SYN+ACK. Mainly response packets.
This means that the IPs are not always in the expected order:
443:1.1.1.1 -> 192.168.1.123:12345
which causes sometimes not to obtain the process of the connection,
because the connection in the system appears as
12345:192.168.1.123 -> 1.1.1.1:443
Intercepting packets with *only* the SYN flag set seems to resolve
this problem.
- Send errors to the server (GUI) if there's any error when reloading
the system fw rules (far from being perfect/optimal, needs a
rewrite).
- Don't load the configuration after saving it, let the watcher reload
it on write change to avoid double reload/duplicated errors.
- Previously we only supported multiple ICMP types on the same rule
by adding multiple keys:
Key: type
Value: echo-request
Key: type
Value: echo-reply
Now it's possible to specify them using ',':
Key: type
Value: echo-request,echo-reply
- Validate ICMP types before adding them.
Now you can add rules to allow multiple protocols.
For example you can add a rule to allow dport/sport for both TCP
and UDP.
There're two options to allow a port:
Statement {
Name: tcp
Values:
Key: dport
Value: 1234
}
Statement {
Name: meta
Values:
Key: l4proto
Value: tcp,udp
Key: dport
Value: 1234
}
Closes#951.
The DNS rule to intercept DNS responses must always be at the top of
the (input-filter) rules, otherwise we won't receive DNS resolutions.
Adding, removing or changing system fw rules was removing the rule from 1st
position.
Another approach to this problem could be to remove&&add only the dns rule,
instead of disable-enable interception+rules monitor.
On some kernels (4.19), adding the interception rule to the
inet-mangle-output chain failed.
According to the nftables wiki, the mangle-output chain have (must?) to
be of type Route:
"route type: ... mangle table ... for the output hook (for other
hooks use type filter instead)."
https://wiki.nftables.org/wiki-nftables/index.php/Configuring_chains#Base_chain_types
So if we fail adding the interception rule, we retry it with type Filter
instead of Route.
Related: #781 , ced9a24
This is part of latest commit ced9a24933
- When reusing a chain, configure the new policy.
- Don't backup existing rules when reloading the configuration.
On some systems after disabling and enabling the interception, the
DNS rule was not being re-added, with error: chain already exists.
When the interception is disabled, we firstly remove the rules,
and secondly if the chain and table is empty we try to remove them
as well.
However, on some systems even if "nft list ruleset" didn't list the
chains to hold our rules, it failed with error "already exists".
Before adding a chain now we check if it exists, and if it does we
reuse it.
Closes#781
- Allow to use commas to define conntrack states:
("related,established")
- Remove profile's rules when switching from Deny to Allow.
- Fixed checking duplicated rules.
Closes#778