previously, we called the restore method from the startup when there was an unclean shutdown. But it never had the state keys to clean since they are stored in memory
this change addresses the issue by falling back to default values when restoring the host's DNS
On macOS use the recommended settings for providing split DNS. As per
the docs an empty string will force the configuration to be the default.
In order to to support split DNS an additional service config is added
for the local server and search domain settings.
see: https://developer.apple.com/documentation/devicemanagement/vpn/dns
* Add faster availability DNS probe and update test domain to .com
- Count success queries and compare it before doing after network map probes.
- Reduce the first dns probe to 500ms
- Updated test domain with com instead of . due to Palo alto DNS proxy server issues
* use fqdn
* Update client/internal/dns/upstream.go
Co-authored-by: Viktor Liu <17948409+lixmal@users.noreply.github.com>
---------
Co-authored-by: Viktor Liu <17948409+lixmal@users.noreply.github.com>
* Enable network monitoring by default for Windows and Darwin
* Enable network monitoring by default on Windows and macOS
* fix merge
* Prevent updating config if no changes
* compile client under freebsd (#1620)
Compile netbird client under freebsd and now support netstack and userspace modes.
Refactoring linux specific code to share same code with FreeBSD, move to *_unix.go files.
Not implemented yet:
Kernel mode not supported
DNS probably does not work yet
Routing also probably does not work yet
SSH support did not tested yet
Lack of test environment for freebsd (dedicated VM for github runners under FreeBSD required)
Lack of tests for freebsd specific code
info reporting need to review and also implement, for example OS reported as GENERIC instead of FreeBSD (lack of FreeBSD icon in management interface)
Lack of proper client setup under FreeBSD
Lack of FreeBSD port/package
* Add DNS routes (#1943)
Given domains are resolved periodically and resolved IPs are replaced with the new ones. Unless the flag keep_route is set to true, then only new ones are added.
This option is helpful if there are long-running connections that might still point to old IP addresses from changed DNS records.
* Add process posture check (#1693)
Introduces a process posture check to validate the existence and active status of specific binaries on peer systems. The check ensures that files are present at specified paths, and that corresponding processes are running. This check supports Linux, Windows, and macOS systems.
Co-authored-by: Evgenii <mail@skillcoder.com>
Co-authored-by: Pascal Fischer <pascal@netbird.io>
Co-authored-by: Zoltan Papp <zoltan.pmail@gmail.com>
Co-authored-by: Viktor Liu <17948409+lixmal@users.noreply.github.com>
Co-authored-by: Bethuel Mmbaga <bethuelmbaga12@gmail.com>
* Removejsonfile' from test matrix in workflows
* Remove sqlite to json migration command
* Refactor store engine implementation to remove JSON file store support
The codebase has been refactored to remove support for JSON file store storage engine, with SQLite serving as the default store engine. New functions have been added to handle unsupported store engines and to migrate data from file store to SQLite.
* Remove 'downCmd' from migration commands
* Refactoring
* Add sqlite cleanup
* Remove comment
Because the connWorker are operating with the e.peerConns list we must ensure all workers exited before we modify the content of the e.peerConns list.
If we do not do that the engine will start new connWorkers for the exists ones, and they start connection for the same peers in parallel.
* migrate sqlite store to
generic sql store
* fix conflicts
* init postgres store
* Add postgres store tests
* Refactor postgres store engine name
* fix tests
* Run postgres store tests on linux only
* fix tests
* Refactor
* cascade policy rules on policy deletion
* fix tests
* run postgres cases in new db
* close store connection after tests
* refactor
* using testcontainers
* sync go sum
* remove postgres service
* remove store cleanup
* go mod tidy
* remove env
* use postgres as engine and initialize test store with testcontainer
---------
Co-authored-by: Maycon Santos <mlsmaycon@gmail.com>
* Enable release workflow on PR and upload binaries
add GetSystem32Command to validate if a command is in the path
it will fall back to the full system32, assuming the OS driver is C
---------
Co-authored-by: Maycon Santos <mlsmaycon@gmail.com>
* config.go: pull unified Config.apply() out of createNewConfig() and update()
as a bonus it ensures returned Config object doesn't have any configuration
values missing
* Add status anonymization
* Add OS/arch to the status command
* Use human-friendly last-update status messages
* Add debug bundle command to collect (anonymized) logs
* Add debug log level command
* And debug for a certain time span command
Now that we have the latency between peers available we can use this data to consider when choosing the best route. This way the route with the routing peer with the lower latency will be preferred over others with the same target network.
* Add Linux legacy routing if ip rule functionality is not available
* Ignore exclusion route errors if host has no route
* Exclude iOS from route manager
* Also retrieve IPv6 routes
* Ignore loopback addresses not being in the main table
* Ignore "not supported" errors on cleanup
* Fix regression in ListenUDP not using fwmarks
All routes are now installed in a custom netbird routing table.
Management and wireguard traffic is now marked with a custom fwmark.
When the mark is present the traffic is routed via the main routing table, bypassing the VPN.
When the mark is absent the traffic is routed via the netbird routing table, if:
- there's no match in the main routing table
- it would match the default route in the routing table
IPv6 traffic is blocked when a default route IPv4 route is configured to avoid leakage.
* adding peer healthcheck
* generate proto file
* fix return in udp mux and replace with continue
* use ice agent for latency checks
* fix status output
* remove some logs
* fix status test
* revert bind and ebpf code
* fix error handling on binding response callback
* extend error handling on binding response callback
---------
Co-authored-by: Maycon Santos <mlsmaycon@gmail.com>
* Fix using wrong array index in log to avoid potential panic
* Increase gRPC connection timeout and add the timeout resolv.conf option
This makes sure the dns client is able to failover to a second
configured nameserver, if present. That is the case then when using the
dns `file` manager and a resolv.conf file generated for netbird.
* On file backup restore, remove the first NS if it's the netbird NS
* Bump dns mangager discovery message from debug to info to ease debugging
This changes the default behavior for new peers, by requiring the agent to be executed with allow-server-ssh set to true in order for the management configuration to take effect.
With these changes, the command up supports the flag --disable-auto-connect that allows users to disable auto connection on the client after a computer restart or when the daemon restarts.
This PR implements the following posture checks:
* Agent minimum version allowed
* OS minimum version allowed
* Geo-location based on connection IP
For the geo-based location, we rely on GeoLite2 databases which are free IP geolocation databases. MaxMind was tested and we provide a script that easily allows to download of all necessary files, see infrastructure_files/download-geolite2.sh.
The OpenAPI spec should extensively cover the life cycle of current version posture checks.
* Make sure our iOS dialer does not get overwritten
* set dial timeout for both clients on ios
---------
Co-authored-by: Pascal Fischer <pascal@netbird.io>
sends an extra server reflexive candidate to the remote peer with our related port (usually the Wireguard port)
this is useful when a network has an existing port forwarding rule for the Wireguard port and the local peer and avoids creating a 1:1 NAT on the local network.
Ensure we use WG address instead of loopback addresses for eBPF.
- First try to use 53 port
- Try to use 5053 port on WG interface for eBPF
- Try to use 5053 on WG interface or loopback interface
