-
-
Notifications
You must be signed in to change notification settings - Fork 512
FAQs
Yes, it does. But only the first time it is used. Once you configure which processes/connections you want to allow/deny, you won't notice that it's running. Really.
In the future, maybe we can add an option to queue events, and allow/deny them from the GUI: applying the default configured action until further interaction from the user.
I tried, but for very fast updates it failed bad on my configuration (failed bad = SIGSEGV), moreover I find Qt5 layout system superior and easier to use.
The UI service is able to use a TCP listener instead of a UNIX socket, that means the UI service itself can be executed on any operating system, while receiving messages from a single local daemon instance or multiple instances from remote computers in the network, therefore DBUS would have made the protocol and logic uselessly GNU/Linux specific.
Be sure that the daemon is running: $ pgrep opensnitchd
If it's not running, you may need to enable and start it:
$ sudo systemctl enable opensnitchd
$ sudo systemctl start opensnitchd.service
Check that the daemon is running: $ pgrep opensnitchd . Status should be "Running".
Click on the Rules -> Nodes -> <node address> , and see if the rules are listed.
Some more info: #988
The common reason is because the eBPF module is not installed or not working.
If you are using the packages from AUR, you need to install the additional package https://aur.archlinux.org/packages/opensnitch-ebpf-module
Another reason could be , that the eBPF module insertion failed or the debugfs is not mounted. In both cases, execute the following command:
$ sudo cat /sys/kernel/debug/tracing/kprobe_event
If the output is empty or it fails, then you can try mounting it: # mount -t debugfs none /sys/kernel/debug then restart opensnitch: sudo service opensnitch restart
If it still doesn't work, you can enable [x] Debug invalid connections under Preferences->Nodes.
We currently (>= v1.6.0-rc.4) intercept new connections (iptables/conntrack state NEW) of TCP, UDP and UDPLITE, SCTP and ICMP protocols, to/from any port.
Your kernel needs some features to be enabled in order eBPF to work: debugfs (or tracefs), kprobes, perf events, ftrace and syscalls (bpf and ftrace).
Since version 1.6.x you can execute the following command to know if your kernel has all the expected features:
opensnitchd -check-requirements
Unfortunately, these kernels are compiled without the mentioned features, so eBPF process monitor method won't be available on these kernels. liquorix does have support for kprobes, but no syscalls tracing. But xanmod doesn't have support for any of the needed features.
For these kernels, the default method to intercept processes will be ProcFS (proc).
We support most of the kernel hardening options. However some of them causes eBPF not to work. We don't know yet (18/11/2022) which is exactly the option that prevent us to work as expected.
When a process establishes a new connection, we first receive the connection information (src/dst IP, src/dst port, but no PID, nor process command line/path). Thus, we try to get who created the connection.
Sometimes we fail to discover the PID of the process, or the path of the PID, thus in these cases if you check this option, a pop-up will appear to allow or deny an "unknown connection".
The daemon option "DefaultAction" "deny" will block ALL traffic (as of version 1.6.0-rc.4) that is intercepted by iptables or nftables and is not answered or configured by the user. If an outgoing connection timeouts while waiting for user action, then it'll apply the default action.
If you suspect that opensnitch is blocking an application and asking you to allow/deny it (for example VPN traffic), enable the option [x] Debug invalid connections from Preferences -> Nodes
If you need to allow this kind of traffic, you can add a rule directly to iptables/nftables:
iptables -t mangle -I OUTPUT -p icmp -j ACCEPT
Read more 👉 https://github.com/evilsocket/opensnitch/wiki/System-rules-legacy
Since version 1.2.0, rules are checked in alphabetical order.
They are evaluated until a rule with a Deny/Reject Action is found, or until a rule with the [x] Priority check marked is found.
So if you want to prioritize some rules over others:
- Name the rule as 000-max-priority, 001-notsomax-priority, 002-less-preiority, not-priority
- Priority field checked (Action: allow)
- OR Action: deny (not need to check the Priority field in these rules)
More info:
- https://github.com/evilsocket/opensnitch/wiki/Rules-examples
- https://github.com/evilsocket/opensnitch/wiki/Rules#best-practices
No. You only allow program A to access the net. Any other program launched by program A will be stopped until you allow or deny it.
See some examples:
- Spotify launching wget: https://github.com/evilsocket/opensnitch/discussions/401
- Vivaldi browser deb package trying to install from the internet additional packages: https://github.com/evilsocket/opensnitch/discussions/742
Read more about best practices: https://github.com/evilsocket/opensnitch/wiki/Rules#best-practices
A process may open other subprocesses, but will it bypass defined application rules? No (see previous FAQ why ^). From the OpenSnitch perspective, it'll just be a new process opening an outbound connection without a rule defined for it, and as such, it'll ask you to allow or deny it.
See examples of malware running on GNU/Linux and how OpenSnitch detects the outbound connections:
https://github.com/evilsocket/opensnitch/discussions/791 https://github.com/evilsocket/opensnitch/discussions/743 https://github.com/evilsocket/opensnitch/discussions/742 https://github.com/evilsocket/opensnitch/discussions/564 https://github.com/evilsocket/opensnitch/discussions/1100
If you create a rule to allow wget or curl system-wide, a malicious process may use of it to download remote files, so it all depends on what rules you define:
https://github.com/evilsocket/opensnitch/wiki/Rules#best-practices
Anyway, nothing is unbreakable. If you know a way to bypass application rules, we'd love to see a detailed example! That'll help us to improve the application.
Do I need to turn off or uninstall other firewalling (firewalld, ufw, gufw) before installing OpenSnitch ? Will the OpenSnitch install or app turn them off automatically ?
No, you don't need to turn off or uninstall other firewalling. OpenSnitch doesn't turn them off, nor delete their rules.
Please help us make this wiki better.
How to submit changes: https://github.com/evilsocket/opensnitch/blob/wiki/README.md
- Installation
- Getting started
- Configuration
- Compilation
- GUI translations
- FAQs and common errors
- Examples OpenSnitch in action