This tool can be used to automatically create wireguard tunnels. Using
rpcd a new wireguard interface is created on the server where the client
can connect to.
Wiregurad server automatically installs a user and associated ACL to use
the wireguard-installer-server features. The user is called wginstaller
and so is the password.
Get Usage:
wg-client-installer get_usage --ip 127.0.0.1 --user wginstaller
--password wginstaller
Register Interface:
wg-client-installer register --ip 127.0.0.1 --user wginstaller
--password wginstaller --bandwidth 10 --mtu 1400
Signed-off-by: Nick Hainke <vincent@systemli.org>
Not including an A record mapping will cause nsupdate to balk at
CNAME and MX records (and probably SRV as well) because the target
will be unknown at the time of parsing, until the lease gets
activated.
We need these RR's to be in place well before the servers even
come up.
Signed-off-by: Philip Prindeville <philipp@redfish-solutions.com>
Microsoft Windows, Xbox and possibly other operating systems do not
support IGDv2. With IGDv2 enabled, they send a HTTP GET request for
rootDesc.xml and WANIPCn.xml, and then nothing happens. The Microsoft
implementation probably doesn't like the WANIPCn.xml response and
decides UPnP is not available. When miniupnpd is built without IGDv2
support, after the 2 HTTP GET requests, there is a HTTP POST request to
/ctl/IPConn, and miniupnpd configures the port forward as expected.
The runtime option force_igd_desc_v1=yes (UCI: igvd1) does not solve
this problem. It's possible this was enough in earlier miniupnpd
versions, but it does not fix the problem the current version.
Since we are a modern distro, we want to support the latest and
greatest, so we should default to IGDv2 enabled. Introducing a
menuconfig option to disable IGDv2 would only help people who build
their own images, so offer a separate package variant for IGDv1.
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
* major rewrite
* add support for multiple chains
* add mac whitelisting
* add support for multiple ssh daemons in parallel
* add an ipset report engine
* add mail notifications
* add suspend/resume functions
* add a cron wrapper to set an ipset related auto-timer for
automatic blocklist updates
* add a list wrapper to add/remove blocklist sources
* add 19.x and Turris OS 5.x compatibility code
* sources stored in an external compressed json file
(/etc/banip/banip.sources.gz)
* change Country/ASN download sources (faster/more reliable)
* fix DHCPv6/icmpv6 issues
Signed-off-by: Dirk Brenken <dev@brenken.org>
Fix starting problem:
Starting function should be named 'start_service' instead of 'start_instance'.
Fix reloading problem:
Register reload tigger for uci config itself.
And, xray does not support reload currently, so use legacy restart as reload.
Fixes: 6c9b96352f ("xray-core: add init script")
Signed-off-by: Tianling Shen <cnsztl@project-openwrt.eu.org>
Major changes are:
add "vfs objects = acl_xattr" parameter in configuration.
fix wrong group domain name in lsarpc response.
set to SID_TYPE_UNKNOWN if there is no domain sid in server.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
The iputils build system embeds git tags into the generated binaries
for use by commands like ping -V. Since openwrt packaging is done in
a different repository from the upstream repo, the tags it finds
aren't particularly meaningful, and we get confusing results like
those described at https://github.com/openwrt/packages/issues/13920
This change removes the git tag inspection in favor of the static
version string that's already known to the upstream build system.
Signed-off-by: Noah Meyerhans <frodo@morgul.net>
Drop obsolete patches
- 001-no-tests.patch
- 002-fix-cross-compilation.patch
Move several user-executable binaries from /usr/sbin to /usr/bin per
upstream.
Signed-off-by: Noah Meyerhans <frodo@morgul.net>
Current implementation of socat's init service doesn't allow to run more
complex configurations. As an example there's no possibility to execute
following command:
socat TCP-LISTEN:8080,fork,reuseaddr,bind=192.168.1.1 \
EXEC:"/sbin/ip netns exec somenetns socat STDIO TCP:10.0.0.1:80"
In such command the first line is argv[1] and the second line is
argv[2]. SocatOptions config option is a string. As as a consequence of
this each word will be passed as a separate argv element. Socat won't be
able to parse arguments correctly.
In order to mitigate this issue, we can also accept SocatOptions as a
list of strings. Following config file will work correctly:
config socat 'tunnel_8080_into_somenetns'
option enable '1'
list SocatOptions 'TCP-LISTEN:8080,fork,reuseaddr,bind=192.168.1.1'
list SocatOptions 'EXEC:"/sbin/ip netns exec somenetns socat STDIO TCP:10.0.0.1:80"'
While we're at it, pass stdout and stderr into logread.
Signed-off-by: Dobroslaw Kijowski <dobo90@gmail.com>
* The default local-adress makes Netopeer2-server listen on ipv4 only.
We change it to :: in order to listen on ipv6 as well as ipv4.
Signed-off-by: Jakov Smolic <jakov.smolic@sartura.hr>
* fix for possible exploit #13758
* sanetize all external template/config inputs
* fix some shellcheck warnings
Signed-off-by: Andy Walsh <andy.walsh44+github@gmail.com>
I checked the hostname for existing DNS A and AAAA entries and these
ones didn't have an entry.
Signed-off-by: Gerald Hansen <gerald.hansen@cloud.ionos.com>
As the default uclient-fetch doesn't support authentication header
and the ddns provider myonlineportal.net support also username and
passwort as url parameter this can be changed.
Signed-off-by: Gerald Hansen <gerald.hansen@cloud.ionos.com>
add eoip package,this can create ethernet
tunnels compatible with Mikrotik EoIP tunnel.
At current moment it is easiest way
to create stateless tunnel with Mikrotik.
Signed-off-by: Bogdan Shatik <bogdikxxx@mail.ru>
In IPv4 the default route can be written as
0.0.0.0/0
In IPv6 the default route can be written as
::/0
If u try
owipcalc 0.0.0.0/0 contains 1.1.1.1
or
owipcalc ::/0 contains ::1
owipcalc will respond with 0 meaning that the "default prefixes" do not
contain the routes.
That is why we check now for 0 prefix.
Furthermore, if the prefix is 0, i will be 16. We will access a negative
array entry in the line:
uint8_t net1 = x->s6_addr[15-i] & m;
Divide by % 16 to prevent i becoming 16:
uint8_t i = ((128 - a->prefix) / 8) % 16;
Signed-off-by: Nick Hainke <vincent@systemli.org>
This is a helpful utility, but it does not have any dependencies
in base repository. Move it to packages feed.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
This is a helpful utility, but it does not have any dependencies
in base repository. Move it to packages feed.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>