Import changes made upstream

docs/caveats.md

@@ -350,7 +350,7 @@ You can change the license file parameters within a node definition or with **de
 
 * *netlab* supports only vJunosEvolved releases that do not require external PFE- and RPIO links. The first vJunosEvolved release implementing internal PFE- and RPIO links is the release 23.2R1-S1.8.
 * vJunosEvolved software is supposed to be run on an Intel CPU. _netlab_ implements the hacks [suggested by Juniper to run vJunos on an AMD CPU](https://www.juniper.net/documentation/us/en/software/nce/nce-510-virtual-switches-mist-cloud-managed/amd-cpu_unofficial_tweaks.html), but please note that this is not supported.
-* _netlab_ can start virtual machines running vJunosEvolved release 23.2 (release 24.1 and later requires UEFI BIOS). Use the vrnetlab-generated container as a workaround if you want to run vJunosEvolved release 24.1 or later.
+* _netlab_ can start virtual machines running vJunosEvolved release 24.1+ (which requires UEFI BIOS - previous releases do not work with UEFI BIOS). Use the vrnetlab-generated container as a workaround if you want to run older vJunosEvolved releases.
 * _netlab_ supports the [vrnetlab-generated](https://containerlab.dev/manual/kinds/vr-vjunosevolved/) vJunosEvolved container, not the cJunosEvolved container. The differences between the two make it impossible to run both container versions as the same virtual device, and cJunosEvolved offers no significant advantages at the moment.
 * The vJunosEvolved container runs on Intel and AMD CPU.
 * The virtual MAC address of the anycast gateway is ignored. _netlab_, therefore, does not support the anycast gateway on vPTX.

docs/module/routing_protocols.md

@@ -37,10 +37,14 @@ nodes:
 (routing_af)=
 ## Address Families
 
-Configuration modules for all IGP routing protocols that support multiple address families (IS-IS, EIGRP) or multiple protocol instances (OSPFv2, OSPFv3) support **_protocol_.af** global- or node-level module parameter. The **af** parameter can be a list- or a dictionary of address families.
+Configuration modules for all IGP routing protocols that support multiple address families (IS-IS, EIGRP) or multiple protocol instances (OSPFv2, OSPFv3) support **_protocol_.af** global-, node-, or VRF-level module parameter. The **af** parameter can be a list or a dictionary of address families.
 
 The default value of the **af** parameter is set based on address families configured on loopback- or physical interfaces -- an address family is enabled within an IGP configuration on a device if at least one interface on that device has an IP address from that address family.
 
+```{warning}
+If you decide to set a **‌_protocol_.af** parameter, you MUST specify which address families should be active. Using the **af** parameter just to set an unwanted address family to *False* without enabling any other address family will result in no address families being used.
+```
+
 **Example:**
 
 ```
@@ -61,7 +65,7 @@ nodes:
   r5:
 ```
 
-* IS-IS address families are enabled based on configured **isis.af** parameters
+* IS-IS address families are enabled based on the configured **isis.af** parameters
 * **ospf.af** parameter is not set -- OSPFv2 or OSPFv3 are enabled based on address families used on individual nodes
 
 The following IS-IS address families are configured on individual routers:

docs/module/vrf.md

@@ -107,6 +107,8 @@ You can also set these parameters to influence routing protocols within a VRF.
 * **ospf.area** -- the default OSPF area for the VRF OSPF process (default: node **ospf.area**). It is configured on the VRF loopback interfaces.
 * **bgp.router_id** -- per-VRF BGP router ID. You have to set this parameter if you want to configure inter-VRF EBGP sessions between interfaces of the same device.[^ELB]
 * **ospf.router_id** -- per-VRF OSPF router ID. You can use this parameter for the same reasons as **bgp.router_id** or if you want consistent OSPF router IDs on Cisco IOS.
+* **_igp_.af** -- specify the [address families](routing_af) you want to use with **isis**, **ospf**, or **rip** VRF instance.
+* You can also specify OSPF route redistribution (**import**), **default** route origination, and default OSPF **password** and **timers**.
 
 [^ELB]: That's how some people implement inter-VRF route leaking. You don't want to know the details ;)
 

netsim/defaults/addressing.yml

@@ -12,7 +12,7 @@ p2p:
 mgmt:
   ipv4: 192.168.121.0/24
   start: 100
-  mac: 08-4F-A9-00-00-00
+  mac: CA-FE-00-00-00-00
 l2only:
 vrf_loopback:
   ipv4: 10.2.0.0/24

netsim/devices/cat8000v.yml

@@ -4,7 +4,7 @@ parent: csr
 virtualbox:
   image:
 clab:
-  image: vrnetlab/vr-c8000v:17.13.01a
+  image: vrnetlab/cisco_c8000v:17.13.01a
   node:
     kind: cisco_c8000v
   interface.name: eth{ifindex-1}

netsim/devices/iosvl2.yml

@@ -31,7 +31,7 @@ clab:
     ansible_ssh_pass: admin
     ansible_user: admin
     netlab_check_retries: 50
-  image: vrnetlab/cisco_viosl2:15.2.2020
+  image: vrnetlab/cisco_viosl2:15.2
   node:
     kind: linux
   interface.name: eth{ifindex}

netsim/devices/srlinux.yml

@@ -25,7 +25,9 @@ clab:
   image: ghcr.io/nokia/srlinux:25.3.3
   node:
     kind: srl
-    type: ixrd2
+    type: ixr-d2
+    config_templates:
+      hosts: /etc/hosts
   interface:
     name: e1-{ifindex}
   group_vars:

netsim/devices/vptx.yml

@@ -24,7 +24,7 @@ libvirt:
   create_template: vptx.xml.j2
 
 clab:
-  image: vrnetlab/juniper_vjunosevolved:23.4R2-S2.1
+  image: vrnetlab/juniper_vjunosevolved:25.2R1.8-EVO
   build: https://containerlab.dev/manual/kinds/vr-vjunosevolved/
   mtu: 1500
   node:

netsim/install/libvirt/vptx.xml.j2

@@ -63,6 +63,8 @@
     <ns0:arg value="type=0,vendor=Bochs,version=Bochs"/>
     <ns0:arg value="-smbios"/>
     <ns0:arg value="type=3,manufacturer=Bochs"/>
+    <ns0:arg value="-bios"/>
+    <ns0:arg value="/usr/share/qemu/OVMF.fd"/>
     <ns0:arg value="-smbios"/>
     <ns0:arg value="type=1,manufacturer=Bochs,product=Bochs,serial=chassis_no=0:slot=0:type=1:assembly_id=0x0D20:platform=251:master=0:channelized=no"/>
   </ns0:commandline>

netsim/modules/_routing.py

@@ -25,38 +25,38 @@
 # * If the address families are not set, calculate them based on interface address families
 # * Otherwise parse and validate the AF attribute
 #
-def routing_af(node: Box, proto: str, features: typing.Optional[Box] = None) -> None:
-  if 'af' in node[proto]:               # Is the AF attribute set for the routing protocol?
-    if isinstance(node[proto].af,list): # Turn a list of address families into a dictionary
-      node[proto].af = { af: True for af in node[proto].af }
-
-    if node[proto].af is None:
-      node[proto].pop('af',None)
-    else:
-      if not isinstance(node[proto].af,dict):
-        log.error(
-          f'af attribute for {proto} on node {node.name} has to be a list or a dictionary',
-          log.IncorrectValue,
-          proto)
-        return
-
-      for proto_af in node[proto].af.keys():
-        if not proto_af in ('ipv4','ipv6'):
-          log.error(
-            f'Routing protocol address family has to be ipv4 and/or ipv6: {proto} on {node.name}',
-            log.IncorrectValue,
-            proto)
+# Please note that this function could be called with node data (for global routing instance)
+# or VRF data (acting as node data) for VRF routing instances
+#
+def routing_af(
+      node: Box,                                      # Could also be VRF data
+      proto: str,                                     # IGP we're checking
+      n_name: typing.Optional[str] = None,            # Name of the object (default: node name)
+      is_vrf: bool = False,                           # Are we dealing with a VRF?
+      features: typing.Optional[Box] = None) -> None: # And finally, the device features
+
+  if n_name is None:
+    n_name = 'node {node.name}'
+  if 'af' in node[proto] and node[proto].af is None:
+    node[proto].pop('af',None)
 
   if not 'af' in node[proto]:                         # No configured AF attribute, calculate it
-    for af in ['ipv4','ipv6']:
-      if af in node.get('loopback',{}):               # Address family enabled on loopback?
-        node[proto].af[af] = True                     # ... we need it in the routing protocol
-        continue
+    for af in ['ipv4','ipv6']:                        # Process the loopback interface
+      lb_data = node.get('loopback',{})               # Fetch the loopback interface info
+      if isinstance(lb_data,Box):                     # ... watch out for bool values
+        if af in lb_data:                             # Address family enabled on loopback?
+          node[proto].af[af] = True                   # ... we need it in the routing protocol
+          continue
+
+      # Scan all interfaces -- they could be in node.interfaces or vrf_data.proto.interfaces
+      if_list = node[proto].get('interfaces',[]) if is_vrf else node.get('interfaces',[])
+      for l in if_list:
+        if 'vrf' in l and not is_vrf:                 # Skip VRF interfaces when processing global routing instance
+          continue                                    # VRF instances will only have correct VRF interfaces anyway
 
-      for l in node.get('interfaces',[]):             # Scan all interfaces
-        if proto in l and (af in l or af in l.get('dhcp.client',{})) \
-           and not 'vrf' in l:                        # Do we have AF enabled on any global interface?
-          node[proto].af[af] = True                   # Found it - we need it the module
+        # Do we have AF enabled on the global- or VRF interface
+        if proto in l and (af in l or af in l.get('dhcp.client',{})):
+          node[proto].af[af] = True                   # Found the AF - we need it in the routing protocol
           continue
 
   p_features = features.get(proto,{}) if features else {}
@@ -68,7 +68,7 @@ def routing_af(node: Box, proto: str, features: typing.Optional[Box] = None) ->
 
     if af in p_features and p_features[af] is False:
       log.error(
-        f'Device {node.device} (node {node.name}) cannot run {proto} on {af}',
+        f'{n_name} cannot run {proto} on {af}',
         log.IncorrectValue,
         proto)
 
@@ -297,6 +297,7 @@ def build_vrf_interface_list(
       vrf_unnumbered_check: bool = True) -> None:                     # Check the source IP VRF for IPv4 unnumbereds
 
   unnum_err_list = data.get_empty_box()
+  features = devices.get_device_features(node,topology.defaults)
 
   for l in node.interfaces:
     if proto not in l:                                                # Not running the protocol on the interface?
@@ -346,7 +347,9 @@ def build_vrf_interface_list(
         f"VRF {vname} on {node.name} has unnumbered interface(s) running {proto} without a VRF loopback",
         category=log.MissingDependency,
         module="vrf",
-        more_data=f"Used on interface(s) {','.join(unnum_err_list[vname])}")      
+        more_data=f"Used on interface(s) {','.join(unnum_err_list[vname])}")
+
+    routing_af(vdata,proto,f'VRF {vname} on {node.name}',is_vrf=True,features=features)
 
 #
 # remove_unaddressed_intf -- remove all interfaces without IPv4 or IPv6 address from IGP
@@ -431,6 +434,7 @@ def check_vrf_protocol_support(
     if not proto in v_data:                       # Are we running the target protocol in the VRF?
       continue
     if af and not af in v_data.af:                # Does the VRF use the target address family?
+      v_data[proto].af.pop(af,None)               # Make sure the AF is not set in VRF routing protocol data
       continue
 
     # We found the protocol to check in the current VRF, time to check device features
@@ -443,9 +447,6 @@ def check_vrf_protocol_support(
         module=proto)
       return
 
-    if af:
-      v_data[proto].af[af] = True                 # Add the AF to VRF routing protocol data
-
 """
 check_intf_support -- check device support for optional interface parameters
 """
@@ -821,7 +822,7 @@ def igp_post_transform(
   else:
     remove_vrf_interfaces(node,proto)
 
-  routing_af(node,proto,features)
+  routing_af(node,proto,f'node {node.name}',features=features)
   if vrf_aware:
     remove_vrf_routing_blocks(node,proto)
 

netsim/modules/isis.yml

@@ -28,6 +28,9 @@ attributes:
     instance:
     net: net
     type:
+  vrf:
+    active: bool
+    af: { copy: global }
   link:
     metric: { type: int, min_value: 1, max_value: 16777215 }
     cost: { type: int, min_value: 1, max_value: 16777215 }

netsim/modules/ospf.yml

@@ -54,6 +54,7 @@ attributes:
   vrf_copy: [ area, router_id, reference_bandwidth ]
   vrf:
     active: bool
+    af: { copy: global }
     area: { copy: global }
     import: _r_import
     default: { copy: node }

netsim/modules/ripv2.yml

@@ -27,6 +27,7 @@ attributes:
     timers:
       copy: global
   vrf:
+    af: { copy: global }
     active: bool
     timers:
       copy: global

netsim/modules/srv6.py

@@ -107,7 +107,7 @@ def node_pre_transform(self, node: Box, topology: Box) -> None:
        node.srv6.locator = locator
     locator_net = ipaddress.IPv6Network(locator)
     if node.get('srv6.allocate_loopback'):                   # Auto-assign a loopback from locator range
-      first_host = next(locator_net.hosts())                 # Use first usable address
+      first_host = next(iter(locator_net.hosts()))           # Use first usable address
       prefix6 = topology.pools['loopback'].get('prefix6',64) # Use loopback.prefix6, default /64
       node.loopback.ipv6 = ipaddress.IPv6Interface((first_host, prefix6)).with_prefixlen
 

netsim/templates/provider/clab/srlinux/hosts-common.j2

@@ -0,0 +1 @@
+../../../../ansible/templates/initial/linux/hosts-common.j2

netsim/templates/provider/clab/srlinux/hosts.j2

@@ -0,0 +1,15 @@
+#
+# Created by netlab (containerlab Linux host)
+#
+127.0.0.1	localhost
+::1	localhost ip6-localhost ip6-loopback
+fe00::0	ip6-localnet
+ff00::0	ip6-mcastprefix
+ff02::1	ip6-allnodes
+ff02::2	ip6-allrouters
+#
+{% if _hosts_entries is defined %}
+{{ _hosts_entries }}
+{% else %}
+{% include 'hosts-common.j2' %}
+{% endif %}

netsim/templates/provider/libvirt/vptx-domain.j2

@@ -12,14 +12,16 @@
       domain.driver = "kvm"
       domain.nic_model_type = "virtio"
       domain.graphics_type = "none"
+      domain.video_type = "none"
       domain.nested = true
-      domain.cpu_mode = "custom"
-      domain.cpu_model = "qemu64"
-{% if 'amd' in defaults.processor %}
-      domain.cpu_feature :name => 'svm', :policy => 'require'
-{% else %}
-      domain.cpu_feature :name => 'vmx', :policy => 'require'
-{% endif %}
+
+{# -- add if condition here for UEFI (rel >24) #}
+      domain.loader = "/usr/share/OVMF/OVMF_CODE_4M.fd"
+      domain.nvram = "/var/lib/libvirt/qemu/nvram/{{ name }}.fd"
+
+      domain.machine_type = "q35"
+      domain.cpu_mode = "host-passthrough"
+
       domain.sysinfo = {
         "bios": {
           "vendor": "Bochs",
@@ -34,8 +36,4 @@
           "manufacturer": "Bochs"
         }
       }
-
-      {% if "amd" in defaults.processor|lower %}
-      domain.cpu_mode = "custom"
-      {% endif %}
     end

tests/errors/dup-mgmt-addr.log

@@ -4,5 +4,5 @@ MissingValue in nodes: Node r6 does not have a usable management IP addresss
 IncorrectValue in nodes: Duplicate management ipv4 address 192.168.121.101 on r2 and r1
 IncorrectValue in nodes: Duplicate management ipv4 address 192.168.121.105 on r7 and r5
 IncorrectValue in nodes: Duplicate management ipv6 address 2001:db8:cafe::1 on r4 and r3
-IncorrectValue in nodes: Duplicate management mac address 08:4f:a9:02:00:00 on r3 and r2
+IncorrectValue in nodes: Duplicate management mac address ca:fe:00:02:00:00 on r3 and r2
 Fatal error in netlab: Cannot proceed beyond this point due to errors, exiting

tests/errors/dup-mgmt-addr.yml

@@ -9,15 +9,15 @@ ospf.timers.hello: 1
 nodes:
   r1:
     mgmt:
-      mac: 0A:4F:A9:01:00:00
+      mac: CA:FE:00:01:00:00
 
   r2:
     mgmt:
       ipv4: 192.168.121.101
 
   r3:
     mgmt:
-      mac: 08:4F:A9:02:00:00
+      mac: CA:FE:00:02:00:00
       ipv6: 2001:db8:cafe::1
 
   r4:

tests/topology/expected/6pe.yml

@@ -185,7 +185,7 @@ nodes:
     mgmt:
       ifname: GigabitEthernet0/0
       ipv4: 192.168.121.104
-      mac: 08:4f:a9:04:00:00
+      mac: ca:fe:00:04:00:00
     module:
     - bgp
     name: ce1
@@ -248,7 +248,7 @@ nodes:
     mgmt:
       ifname: GigabitEthernet0/0
       ipv4: 192.168.121.105
-      mac: 08:4f:a9:05:00:00
+      mac: ca:fe:00:05:00:00
     module:
     - bgp
     name: ce2
@@ -317,7 +317,7 @@ nodes:
     mgmt:
       ifname: GigabitEthernet0/0
       ipv4: 192.168.121.103
-      mac: 08:4f:a9:03:00:00
+      mac: ca:fe:00:03:00:00
     module:
     - isis
     - mpls
@@ -429,7 +429,7 @@ nodes:
     mgmt:
       ifname: GigabitEthernet0/0
       ipv4: 192.168.121.101
-      mac: 08:4f:a9:01:00:00
+      mac: ca:fe:00:01:00:00
     module:
     - isis
     - bgp
@@ -545,7 +545,7 @@ nodes:
     mgmt:
       ifname: GigabitEthernet0/0
       ipv4: 192.168.121.102
-      mac: 08:4f:a9:02:00:00
+      mac: ca:fe:00:02:00:00
     module:
     - isis
     - bgp