Age | Commit message (Collapse) | Author |
|
T5779: conntrack: Apply fixes to <set system conntrack timeout custom>
|
|
accel-ppp: T5688: Fixed migration script for pppoe-server
|
|
Fixed migration script for pppoe-server
|
|
Remove what was not working on 1.3, migrate what was working to new syntax and extend feature for ipv6.
|
|
accel-ppp: T5688: Standardized pool configuration in accel-ppp
|
|
Standardized pool configuration for all accel-ppp services.
1. Only named pools are used now.
2. Allows all services to use range in x.x.x.x/mask
and x.x.x.x-x.x.x.y format
3. next-pool can be used in all services
2. Allows to use in ipoe gw-ip-address without pool configuration
which allows to use Fraimed-IP-Address attribute by radius.
3. Default pool name should be explicidly configured
with default-pool.
4. In ipoe netmask and range subnet can be different.
|
|
ddclient: T5791: Simplify and fix migration script for dynamic dns
|
|
Mark 'dns dynamic name' as tag node to avoid unexpected nesting.
|
|
vti: T5769: restore interface settings on down -> up event
|
|
ddclient: T5791: Update dynamic dns configuration path for consistency [followup]
|
|
|
|
On VTI interface link down the link-local IPv6 address is removed. As soon as
the IPSec tunnel is online again, vti-up-down helper is called which only places
the interface in up state using iproute2 command
sudo ip link set vti0 up
This does not restore the IPv6 LL address. Instead use vyos.ifconfig to properly
re-initialize the VTI interface using the generic update() method.
|
|
T5796:add/fixed OCSERV HTTP security headers
|
|
|
|
mdns: T5793: Cleanup avahi-daemon configuration in `/etc` [followup]
|
|
`/etc/avahi` technically can be deleted since we operate with
avahi-daemon configuration in `/run/avahi-daemon`.
But we still need to keep `/etc/avahi/services` because avahi-daemon
`chroot` to that location at startup. This is setup at build time via
`AVAHI_CONFIG_DIR` and there is no way to change it at runtime.
|
|
mdns: T5793: Cleanup avahi-daemon configuration in `/etc`
|
|
`/etc/avahi` can be deleted since we operate with avahi-daemon
configuration in `/run/avahi-daemon`.
|
|
ddclient: T5791: Update dynamic dns configuration path for consistency
|
|
policy: T4704: Allowed to set metric (MED) to (+/-)rtt
|
|
|
|
|
|
As followup to interface definition change, remove XML snippets that
aren't used anymore. They were there because they were 'include'-ed
multiple times in the interface definition `dynamic-dns.xml.in`. Since
that's not the case anymore, they can be removed.
|
|
Modify the configuration path to be consistent with the usual dialects
of VyoS configuration (wireguard, dns, firewall, etc.)
This would also shorten the configuration path and have a unified
treatment for RFC2136-based updates and other 'web-service' based updates.
While at it, add support for per-service web-options. This would allow
for probing different external URLs on a per-service basis.
|
|
Allowed to set metric (MED) to (+/-)rtt in the route-map.
|
|
image-tools: T5789: copy ssh host keys on image update
|
|
|
|
http-api: T5782: use single config-mode script for https and http-api
|
|
image-tools: T5751: allow non-interactive add/delete image
|
|
|
|
|
|
|
|
|
|
vyos.utils: T5749: fix get_vrf_members() call to iproute2
|
|
The iproute2 master argument is used for both a VRF and a bridge device. Using
this in the VRF context would retrieve and report back the wrong interfaces:
Old implementation:
===================
>>> from vyos.utils.network import get_vrf_members
>>> get_vrf_members('br1')
['eth1', 'eth2', 'vxlan1']
>>> get_vrf_members('black')
['br1.3002', 'br1.4000', 'pim6reg10200']
The new implementation:
=======================
>>> from vyos.utils.network import get_vrf_members
>>> get_vrf_members('br1')
[]
>>> get_vrf_members('black')
['br1.3002', 'br1.4000', 'pim6reg10200']
|
|
|
|
smoketest: T5783: check for any abnormal daemon termination
|
|
image-tools: T4516: exit grub-update service if running in docker
|
|
|
|
We need to ensure when stressing FRR with the smoketests that no unexpected
crash happens. We simply verify the PID of the individual FRR daemons.
|
|
remote: T5773: Fix for broken config download
|
|
|
|
|
|
http-api: T5768: remove auxiliary http-api.conf
|
|
|
|
https api: T5772: check if keys are configured unless PAM auth is enabled for GraphQL
|
|
T5637: firewall: extend rule for default-action to firewall bridge
|
|
pppoe: T5630: make MRU default to MTU if unspecified
|
|
to be able to catch logs using separte rule for default-action
|
|
This fixes the implementation in e062a8c11 ("pppoe: T5630: allow to specify MRU
in addition to already configurable MTU") and restores the bahavior that MRU
defaults to MTU if MRU is not explicitly set.
This was the behavior in VyOS 1.3.3 and below before we added ability to define
the MRU value.
|