| Age | Commit message (Collapse) | Author |
|---|
|
|
|
QoS: T4284: re-implementation using XML and Python
|
|
Set fq-codel as the default queuing discipline for the shaper traffic-policy if
queue-type has not been configured.
fq-codel has been the default qdisc for OpenWRT as well as systemd (and thus
various linux distributions) for a decent while now. It has proven itself to be
a highly effective qdisc for fighting bufferbloat whilst maintaining high link
utilization. The combination of HTB + fq-codel has also seen wide deployment
in cases when additional traffic classification and/or rate/burst management
is required.
This change will allow the configuration of a shaper traffic-policy with
suitable defaults, saving the pain of having to specify fq-codel as a
queue-type for every class.
|
|
|
|
|
|
|
|
complete, finalized PR object from GitHub
|
|
|
|
|
|
|
|
T4897: vxlan: Fix setting `source-address` and `source-interface`
|
|
`leaf_node_changed` returns `[]` (empty list) after a leaf node is
added.
e.g. Setting `source-interface` doesn't work on an existing vxlan
interface.
Steps to reproduce:
- Add a vxlan interface without `source-address` or `source-interface` options set:
```
set interfaces vxlan vxlan999 vni 999
set interfaces vxlan vxlan999 remote 192.168.100.100
commit
```
- Then set `source-address` or `source-interface`:
```
set interfaces vxlan vxlan999 source-interface eth0
commit
```
Actual result:
Source address or source-interface are not set:
```
ip -d link show dev vxlan999
76: vxlan999: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1450 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
link/ether 56:08:ba:4d:4e:a8 brd ff:ff:ff:ff:ff:ff promiscuity 0 minmtu 68 maxmtu 65535
vxlan id 999 remote 192.168.100.100 srcport 0 0 dstport 8472 tos inherit ttl 16 ageing 300 udpcsum noudp6zerocsumtx noudp6zerocsumrx addrgenmode none numtxqueues 1 numrxqueues 1 gso_max_size 65536 gso_max_segs 65535
```
Expected result:
```
77: vxlan999: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1450 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
link/ether 9e:05:d9:58:1a:af brd ff:ff:ff:ff:ff:ff promiscuity 0 minmtu 68 maxmtu 65535
vxlan id 999 remote 192.168.100.100 dev eth0 srcport 0 0 dstport 8472 tos inherit ttl 16 ageing 300 udpcsum noudp6zerocsumtx noudp6zerocsumrx addrgenmode none numtxqueues 1 numrxqueues 1 gso_max_size 65536 gso_max_segs 65535
```
All invocations of leaf_node_changed() should be migrated to is_node_changes() if you are only interested in if something changed and don‘t care what exactly changed (content).
|
|
leaf_node_changed()
We only need to use leaf_node_changed() if we are interested in the detailed
change to a CLI node (what was the previous content). If we are only interested
in if a node changed "at all" is_node_changed() should be used.
|
|
We only need to use leaf_node_changed() if we are interested in the detailed
change to a CLI node (what was the previous content). If we are only interested
in if a node changed "at all" is_node_changed() should be used.
|
|
We only need to use leaf_node_changed() if we are interested in the detailed
change to a CLI node (what was the previous content). If we are only interested
in if a node changed "at all" is_node_changed() should be used.
|
|
T4898: Add mtu config option for dummy interfaces
|
|
I use dummy interfaces in a VRF as source-interfaces for VXLAN in order to force VXLAN send underlay UDP traffic through the VRF where the dummy interface resides.
However dummy interface has no mtu option so it always gets an MTU of 1500. This will cause an error when the mtu of dummy is not large enough for the VXLAN traffic.
Adding this option in the config template will solve this.
|
|
|
|
|
|
Instead of hardcoding the default protocol as TCP in the Python script we shall
use the XML based defaultValue approach instead. This also automatically exports
the default to the CLI completion helper.
|
|
If port is specified on the CLI so must be source and destination under
the port node.
|
|
Commit 60c80df4 ("container: T4870: bump package version 0 -> 1 for filesystem
change") introduced a fundamental change in the container storage driver that
required a manual migration step from the user to not loose any data.
This commit removes the manual user interaction and temporary exports the
container images and re-imports them after the filesystem got changed. The only
things that get lost are orphaned container images no longer referenced by the
CLI - thats an adequate trade-off as those images can always be re-added to the
system.
|
|
overlay2 is the preferred storage driver for all currently supported Linux
distributions, and requires no extra configuration.
|
|
firewall: T2199: Add mac-address match to destination side
|
|
|
|
|
|
move from vfs to overlay driver
The following pre iage upgrade script must be executed to have containers after
the reboot:
for pod in $(cli-shell-api listActiveNodes container name); do
systemctl stop vyos-container-${pod//\'}.service
done
sed -i 's/vfs/overlay/g' /etc/containers/storage.conf /usr/share/vyos/templates/container/storage.conf.j2
rm -rf /usr/lib/live/mount/persistence/container/storage/libpod
for pod in $(cli-shell-api listActiveNodes container name); do
image=$(cli-shell-api returnActiveValue container name ${pod//\'} image)
podman image pull $image
systemctl start vyos-container-${pod//\'}.service
done
for dir in vfs vfs-containers vfs-images vfs-layers; do
rm -rf /usr/lib/live/mount/persistence/container/storage/$dir
done
|
|
container: T4870: Update podman to use overlay storage driver
|
|
ipsec.py
|
|
|
|
Our python scripts use the shebang logic to set an intepreter - we should
rely on this and not use an external interpreter in front of the helper.
|
|
|
|
|
|
script
Sorting DHCP pools and filtering for state can now be done using the new op-mode
mode scripts in DHCP. This allows us to drop the old helpers show_dhcp.py and
show_dhcpv6.py.
|
|
script
Remaining functionality to filter NAT translations for a given address
got implemented to nat.py - with this cahnge we can drop the old files
show_nat*.py
|
|
T4890: Fixed op_mode show conntrack table ipv4
|
|
Fixed op_mode show conntrack table ipv4
Created check on empty column "mark"
|
|
T4866: rewrite show_interfaces.py show* functions to standardized op-mode
|
|
|
|
|
|
|
|
|
|
- show log router-advert
- monitor log router-advert
|
|
Commit 13071a4a ("T4809: radvd: Allow the use of AdvRASrcAddress") added a new
feature to set the RA source-address. Unfortunately it missed a semicolon.
|
|
|
|
T4886: Firewall and route policy: Add connection-mark feature to vyos.
|
|
T4879: IPsec migration script remote-id for peer name eq address
|
|
Migration for "remote-id" where peer is IPv4 or IPv6 address
was missed
It was only migration if peer starts with "@"
It cause that you must manualy set 'remote-id' to get it working
correctly
replace 'vpn ipsec site-to-site peer 192.0.2.2'
=> 'vpn ipsec site-to-site peer peer_192-0-2-2 authentication remote-id 192.0.2.2'
|
|
Commit cca7ec3e ("T4832: dhcp: Add IPv6-only dhcp option support (RFC 8925)")
extended the smoketests but used int over str when calilng the CLI wrapper.
This led to:
TypeError: sequence item 7: expected str instance, int found
|
|
|
