Age | Commit message (Collapse) | Author |
|
T5706: Add custom systemd udev rules to exclude dynamic interfaces (backport #2436)
|
|
Add custom systemd udev rules to exclude some regular and dynamic
interfaces from "systemd-sysctl" calls.
It fixes high CPU utilization (100%) as we have a lot of calls per
interface for dynamic interfaces like ppp|ipoe|sstp etc.
/lib/systemd/systemd-udevd should not be called for those interfaces
(cherry picked from commit ca9cc86233520eb495c17602bf7a110094c1d8e7)
|
|
If we add a new interface for PPPoe-server we MUST restart the
`accel-ppp@pppoe.service` as `reload` is not implemented for
accel-ppp daemon
Otherwise we have listen interface in the /run/accel-pppd/pppoe.conf
which does not work
(cherry picked from commit ffda9068b22e2d8a6841fcd8cdf62bbe266ea02c)
|
|
T5698 EVPN ESI Multihoming (backport #2416)
|
|
set interfaces bonding bond10 evpn es-df-pref '50'
set interfaces bonding bond10 evpn es-id '10'
set interfaces bonding bond10 evpn es-sys-mac '01:23:45:67:89:ab'
set interfaces bonding bond10 member interface 'eth3'
set interfaces bonding bond10 mode '802.3ad'
(cherry picked from commit 937685608e61151275c4f60c6d00c0154f2ca06d)
|
|
Commit 7f6624f5a6f8bd ("vxlan: T3700: support VLAN tunnel mapping of VLAN aware
bridges") added support for Single VXLAN Device (SVD) containers supported by
the Linux Kernel.
When working with bridge VIFs it turned out that when deleting a VIF all the
VXLAN tunnel mappings got deleted, too. In order to avoid this, if the bridge
has a VXLAN member interface which vlan-to-vni mapping enabled, we add a
dependency that we call VXLAN conf-mode script after messing arround with the
bridge VIFs and re-create tunnel mappings.
(cherry picked from commit fdf7f3a05edbaaf8aeca7e24a9980d5af67dca18)
|
|
|
|
T4726: Remove accel-ppp RADIUS vendor validators (backport #2423)
|
|
T5513: firewall - op-mode command backport
|
|
Changing the public key of a peer (updating the key material) left the old
WireGuard peer in place, as the key removal command used the new key.
WireGuard only supports peer removal based on the configured public-key, by
deleting the entire interface this is the shortcut instead of parsing out all
peers and removing them one by one.
Peer reconfiguration will always come with a short downtime while the WireGuard
interface is recreated.
(cherry picked from commit 2fc8738bc9c2fb6364a22d86079e8635cee91949)
|
|
|
|
The vendor name could contain Uppercase or lowercase symbols and
not rely on the dictionary name but on dictionary value
/ # cat /usr/share/freeradius/dictionary.cisco | grep -i vendor
VENDOR Cisco 9
Another example
VENDOR Alcatel-IPD 6527
This way if we use `vendor=cisco` instead of `vendor=Cisco` it
will not work at all
Delete vendor validators
(cherry picked from commit bbc7cabc6be0d5f8629724e9b0025e425168e1a8)
|
|
firewal, nat and nat66.
(cherry picked from commit 51abbc0f1b2ccf4785cf7f29f1fe6f4af6007ee6)
|
|
In order to minimize the flooding of ARP and ND messages in the VXLAN network,
EVPN includes provisions [1] that allow participating VTEPs to suppress such
messages in case they know the MAC-IP binding and can reply on behalf of the
remote host. In Linux, the above is implemented in the bridge driver using a
per-port option called "neigh_suppress" that was added in kernel version 4.15.
[1] https://www.rfc-editor.org/rfc/rfc7432#section-10
(cherry picked from commit ec9a95502daa88b9632af12524e7cefebf86bab6)
|
|
As we have a bunch of options under "paramteres" already and "external" is
clearly one of them it should be migrated under that node as well.
(cherry picked from commit cc7ba8824a5e9ec818f0bbe7fb85e1713a591527)
|
|
T5643: nat: add interface-groups to nat. Use same cli structure for i… (backport #2355)
|
|
Try to have as few calls to sudo in the op-mode scripts as possible. The XML
definitions can deal with it.
(cherry picked from commit 428dee29d36cc3629990ec41afef887821886834)
|
|
|
|
|
|
The current named for certificates are hardcoded in generated config to:
- ca.pem
- cert.pem.key
- cert.pem
It cause a generated config certificates and certificates itself
are different (test-cert-1.pem and ca.pem)
bind :::8080 v4v6 ssl crt /run/haproxy/test-cert-1.pem
/run/haproxy/ca.pem
It is a bug of initial impelemtation. Fix required correct names
from PKI certificates
(cherry picked from commit 0431f1b32c1fc90de82adea5a7e63dad1416c340)
|
|
Updated spacing.
(cherry picked from commit f39a35338ac967381356f8b9b499ec1d730653fc)
|
|
(cherry picked from commit 5180622cd6c928812a644f427d65acae763c37cc)
|
|
interface-name|interface-group as in firewall.
(cherry picked from commit 2f2c3fa22478c7ba2e116486d655e07df878cdf4)
|
|
(cherry picked from commit fca8cce1c114f28cf2db8a0fe2ed7f8b37ea010c)
|
|
|
|
If ethernet interface is a bond memeber:
1. Allow for changing only specific parameters which are specified
in EthernetIf.get_bond_member_allowed_options function.
2. Added inheritable parameters from bond interface to ethernet
interface which are scpecified
in BondIf.get_inherit_bond_options.
Users can change inheritable options under ethernet interface
but in commit it will be copied from bond interface.
3. All other parameters are denied for changing.
Added migration script. It deletes all denied parameters under
ethernet interface if it is a bond member.
(cherry picked from commit aa0282ceb379df1ab3cc93e4bd019134d37f0d89)
|
|
(cherry picked from commit 719a3622f35a0596ffd8a0bd28c071fdaf930153)
|
|
Currently VyOS VXLAN implementation uses the Linux assigned port 8472 that
predates the IANA assignment. As Most other vendors use the IANA assigned port,
follow this guideline and use the new default port 4789.
Existing configuration not defining an explicit port number will be migrated
to the old default port number of 8472, keeping existing configurations work!
(cherry picked from commit 6db8d3ded19f652b99231be0d705d76b598ac72a)
# Conflicts:
# interface-definitions/include/version/interfaces-version.xml.i
|
|
T5541: firewall: re-add zone-based firewall.
|
|
|
|
(cherry picked from commit cd54195d070e49aa084c325b83a71621a4011c97)
|
|
T4913: migrate wireless scripts to new op-mode style (backport #2373)
|
|
(cherry picked from commit 4c4c2b1f8a58398798f20c252bde80461320d330)
|
|
(cherry picked from commit ed29faeea1354dc2bec544c63e55c1c666e0d900)
|
|
This fixes sending packets to uacctd using a socket.
(cherry picked from commit 7a0af0d00bae9179c89155e4b2e6ce94abb29c05)
|
|
'generate tech-support archive' moved to vyos-1x.
Output of 'show tech-support report' command is added to archive.
The default location of the archive is moved to '/tmp'.
The script is rewritten to Python.
(cherry picked from commit 65911b17340a7894aba973113d83ab43964bbf99)
|
|
T5165: Implement policy local-route source and destination port (backport #2342)
|
|
pmacct daemons have one very important specific - they handle control signals in
the same loop as packets. And packets waiting is blocking operation.
Because of this, when systemctl sends SIGTERM to uacctd, this signal has no
effect until uacct receives at least one packet via nflog. In some cases, this
leads to a 90-second timeout, sending SIGKILL, and improperly finished tasks.
As a result, a working folder is not cleaned properly.
This commit contains several changes to fix service issues:
- add a new nftables table for pmacct with a single rule to get the ability to
send a packet to nflog and unlock uacctd
- remove PID file options from the uacctd and a systemd service file. Systemd
can detect proper PID, and PIDfile is created by uacctd too late, which leads
to extra errors in systemd logs
- KillMode changed to mixed. Without this, SIGTERM is sent to all plugins and
the core process exits with status 1 because it loses connection to plugins too
early. As a result, we have errors in logs, and the systemd service is in a
failed state.
- added logging to uacctd
- systemctl service modified to send packets to specific address during a service
stop which unlocks uacctd and allows systemctl to finish its work properly
(cherry picked from commit e364e9813b6833f6b108e7177ef7ea2d9e7bac33)
|
|
T5489: Change default qdisc from 'fq' to 'fq_codel' (backport #2349)
|
|
Add `policy local-route` source and destination port
set policy local-route rule 23 destination port '222'
set policy local-route rule 23 protocol 'tcp'
set policy local-route rule 23 set table '123'
set policy local-route rule 23 source port '8888'
% ip rule show prio 23
23: from all ipproto tcp sport 8888 dport 222 lookup 123
(cherry picked from commit ff43733074675b94ce4ead83fe63870b6cf953c5)
|
|
(cherry picked from commit 93d2ea7d635c7aa5acf3000654393ea48b7c6405)
|
|
(cherry picked from commit 7d597a6dca15cb592230b349ef7ef565f258cf43)
|
|
(cherry picked from commit ac1bd7c2f69e058f54084decbfe6b6d329df6462)
|
|
(cherry picked from commit e357258e645cf85de0035d4ecfbf99db4dd90f7e)
|
|
(cherry picked from commit 27605426a4ad613f45d36e7db5b1664dc3192981)
|
|
Calling system-login.py with no mounted VyOS config has the negative effect
that the script will not detect any local useraccounts and thus assumes they
all need to be removed from the password backend.
As soon as the VyOS configuration is mounted and the CLI content is processed,
system-login.py get's invoked and re-creates the before deleted user accounts.
As the account names are sorted in alphabetical order, the name <-> UID mapping
can get mixed up during system reboot.
The intention behind calling system-login.py from vyos-router init was to
reset system services (PAM, NSS) back to sane defaults with the defaults
provided via system-login.py. As PAM is already reset in vyos-router startup
script, /etc/nsswitch.conf was the only candidate left.
This is now accomplished by simply creating a standard NSS configuration file
tailored for local system accounts.
This is the second revision after the first change via commit 64d32329958
("login: T5521: home directory owner changed during reboot") got reverted.
(cherry picked from commit 12069d5653034b46a47430353c3867b3678c196f)
|
|
This reverts commit 074870dad33d80e78128736f9e89bdfa1a0e08fd.
|
|
During system startup the system-login.py script is invoked by vyos-router
systemd service. As there is no complete configuration available at this
point in time - and the sole purpose of this call is to reset/re-render
the system NSS/PAM configs back to default - it accidently also deleted the
local useraccounts.
Once the VyOS configuration got mounted, users got recreated in alphabetical
order and thus UIDs flipped and the /home suddenely belonged to a different
account.
This commit prevents any mangling with the local userdatabase during VyOS
bootup phase.
(cherry picked from commit 64d323299586da646ca847e78255ff2cd8464578)
|
|
(cherry picked from commit 646f08fc5a302e08aad90af3fa0ee32e138ee585)
|
|
vyos@vyos:~$ show system login users
Username Type Locked Tty From Last login
---------- ------ -------- ----- ------------- ------------------------
vyos vyos False pts/0 172.16.33.139 Mon Oct 2 20:42:24 2023
(cherry picked from commit 80f08af76db0ccee4d6dc1a99b6d8d90884fa33f)
|