You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The OVN-Kubernetes Container Network Interface (CNI) plugin is the default networking solution for a {microshift-short} node. OVN-Kubernetes is a virtualized network for pods and services that is based on Open Virtual Network (OVN).
9
+
[role="_abstract"]
10
+
You can use the OVN-Kubernetes Network Interface to create and manage network connections for internet-connected nodes.
10
11
11
-
* Default network configuration and connections are applied automatically in {microshift-short} with the `microshift-networking` RPM during installation.
12
-
* A node that uses the OVN-Kubernetes network plugin also runs Open vSwitch (OVS) on the node.
13
-
* OVN-K configures OVS on the node to implement the declared network configuration.
14
-
* Host physical interfaces are not bound by default to the OVN-K gateway bridge, `br-ex`. You can use standard tools on the host for managing the default gateway, such as the Network Manager CLI (`nmcli`).
15
-
* Changing the CNI is not supported on {microshift-short}.
16
-
17
-
Using configuration files or custom scripts, you can configure the following networking settings:
18
-
19
-
* You can use subnet CIDR ranges to allocate IP addresses to pods.
20
-
* You can change the maximum transmission unit (MTU) value.
21
-
* You can configure firewall ingress and egress.
22
-
* You can define network policies in the {microshift-short}, including ingress and egress rules.
23
-
* You can use the {microshift-short} Multus plugin to chain other CNI plugins.
Networking features available with {microshift-short} {product-version} include:
33
-
34
-
* Kubernetes network policy
35
-
* Dynamic node IP
36
-
* Custom gateway interface
37
-
* Second gateway interface
38
-
* Node network on specified host interface
39
-
* Blocking external access to NodePort service on specific host interfaces
40
-
41
-
Networking features not available with {microshift-short} {product-version}:
42
-
43
-
* Egress IP/firewall/QoS: disabled
44
-
* Hybrid networking: not supported
45
-
* IPsec: not supported
46
-
* Hardware offload: not supported
47
-
48
-
[id="microshift-ip-forward_{context}"]
49
-
== IP forward
50
-
The host network `sysctl net.ipv4.ip_forward` kernel parameter is automatically enabled by the `ovnkube-master` container when started. This is required to forward incoming traffic to the CNI. For example, accessing the NodePort service from outside of a node fails if `ip_forward` is disabled.
Bridge mappings allow provider network traffic to reach the physical network. Traffic leaves the provider network and arrives at the `br-int` bridge. A patch port between `br-int` and `br-ex` then allows the traffic to traverse to and from the provider network and the edge network. Kubernetes pods are connected to the `br-int` bridge through virtual ethernet pair: one end of the virtual ethernet pair is attached to the pod namespace, and the other end is attached to the `br-int` bridge.
Understand how provider network traffic reaches the physical network through bridge mappings. The following concepts apply:
11
+
12
+
* Traffic leaves the provider network and arrives at the `br-int` bridge.
13
+
* A patch port between `br-int` and `br-ex` then allows the traffic to traverse to and from the provider network and the edge network.
14
+
* Kubernetes pods are connected to the `br-int` bridge through a virtual ethernet pair. One end of the virtual ethernet pair is attached to the pod namespace, and the other end is attached to the `br-int` bridge.
The following table summarizes the status of networking features and capabilities that are either present as defaults, supported for configuration, or not available with the {microshift-short} service:
7
11
8
12
.{microshift-short} networking features and capabilities overview
@@ -30,24 +34,31 @@ The following table summarizes the status of networking features and capabilitie
30
34
31
35
|Egress router|Not available|N/A
32
36
33
-
|Firewall|No^[4]^|Yes
37
+
|Firewall|No|Yes
34
38
35
39
|Hardware offloading|Not available|N/A
36
40
37
41
|Hybrid networking|Not available|N/A
38
42
39
43
|IPsec encryption for intra-cluster communication|Not available|N/A
40
44
41
-
|IPv6|Supported^[5]^|N/A
45
+
|IPv6|Supported|N/A
42
46
43
-
|Ingress router|Yes|Yes^[6]^
47
+
|Ingress router|Yes|Yes
44
48
45
-
|Multiple networks plug-in|Yes|Yes
49
+
|Multiple networks plugin|Yes|Yes
46
50
|===
47
51
48
-
1. If unset, the default value is set to the next immediate subnet after the service network. For example, when the service network is `10.43.0.0/16`, the `advertiseAddress` is set to `10.44.0.0/32`.
49
-
2. You can use the multicast DNS protocol (mDNS) to allow name resolution and service discovery within a Local Area Network (LAN) using multicast exposed on the `5353/UDP` port.
50
-
3. There is no built-in transparent proxying of egress traffic in {microshift-short}. Egress must be manually configured.
51
-
4. Setting up the firewalld service is supported by {op-system-ostree}.
52
-
5. IPv6 is supported in both single-stack and dual-stack networks with the OVN-Kubernetes network plugin. IPv6 can also be used by connecting to other networks with the {microshift-short} Multus CNI plugin.
53
-
6. Configure by using the {microshift-short}`config.yaml` file.
52
+
Additional details about networking capabilities::
53
+
54
+
* `Advertise address`: If unset, the default value is set to the next immediate subnet after the service network. For example, when the service network is `10.43.0.0/16`, the `advertiseAddress` is set to `10.44.0.0/32`.
55
+
56
+
* `Multicast DNS`: You can use the multicast DNS protocol (mDNS) to allow name resolution and service discovery within a Local Area Network (LAN) using multicast exposed on the `5353/UDP` port.
57
+
58
+
* `Network proxies`: There is no built-in transparent proxying of egress traffic in {microshift-short}. Egress must be manually configured.
59
+
60
+
* `Firewall`: Setting up the firewalld service is supported by {op-system-ostree}.
61
+
62
+
* `IPv6`: Is supported in both single-stack and dual-stack networks with the OVN-Kubernetes network plugin. You can also use IPv6 by connecting to other networks with the {microshift-short} Multus CNI plugin.
63
+
64
+
* `Ingress router`: Configure by using the {microshift-short}`config.yaml` file.
The OVN-Kubernetes Container Network Interface (CNI) plugin is the default networking solution for a {microshift-short} node. OVN-Kubernetes is a virtualized network for pods and services that is based on Open Virtual Network (OVN).
11
+
12
+
* Changing the CNI is not supported on {microshift-short}.
13
+
* Default network configuration and connections are applied automatically in {microshift-short} with the `microshift-networking` RPM during installation.
14
+
* A node that uses the OVN-Kubernetes network plugin also runs Open vSwitch (OVS) on the node.
15
+
* OVN-K configures OVS on the node to implement the declared network configuration.
16
+
* Host physical interfaces are not bound by default to the OVN-K gateway bridge, `br-ex`. You can use standard tools on the host for managing the default gateway, such as the Network Manager CLI (`nmcli`).
17
+
18
+
Using configuration files or custom scripts, you can configure the following networking settings:
19
+
20
+
* You can use subnet CIDR ranges to allocate IP addresses to pods.
21
+
* You can change the maximum transmission unit (MTU) value.
22
+
* You can configure firewall ingress and egress.
23
+
* You can define network policies in the {microshift-short}, including ingress and egress rules.
24
+
* You can use the {microshift-short} Multus plugin to chain other CNI plugins.
You must use `ip_forward` to access network connectivity.
11
+
12
+
The host network `sysctl net.ipv4.ip_forward` kernel parameter is automatically enabled by the `ovnkube-master` container when started. This is required to forward incoming traffic to the CNI. For example, accessing the NodePort service from outside of a node fails if `ip_forward` is disabled.
Copy file name to clipboardExpand all lines: modules/microshift-nw-components-svcs.adoc
+10-3Lines changed: 10 additions & 3 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -3,9 +3,16 @@
3
3
// * microshift_networking/microshift-cni.adoc
4
4
5
5
:_mod-docs-content-type: CONCEPT
6
-
[id="microshift-network-comps-svcs_{context}"]
6
+
[id="microshift-nw-components-svcs_{context}"]
7
7
= {microshift-short} networking components and services
8
-
This brief overview describes networking components and their operation in {microshift-short}. The `microshift-networking` RPM is a package that automatically pulls in any networking-related dependencies and systemd services to initialize networking, for example, the `microshift-ovs-init` systemd service.
8
+
9
+
[role="_abstract"]
10
+
Understand networking components and services and their operation in {microshift-short}.
11
+
12
+
[NOTE]
13
+
====
14
+
The `microshift-networking` RPM is a package that automatically pulls in any networking-related dependencies and systemd services to initialize networking, for example, the `microshift-ovs-init` systemd service.
15
+
====
9
16
10
17
NetworkManager::
11
18
NetworkManager is required to set up the initial gateway bridge on the {microshift-short} node. The NetworkManager and `NetworkManager-ovs` RPM packages are installed as dependencies to the `microshift-networking` RPM package, which contains the necessary configuration files. NetworkManager in {microshift-short} uses the `keyfile` plugin and is restarted after installation of the `microshift-networking` RPM package.
@@ -32,4 +39,4 @@ OVN-Kubernetes manifests and startup logic are built into {microshift-short}. Th
32
39
* `/etc/systemd/system/ovsdb-server.service.d/microshift-cpuaffinity.conf` for `ovs-server.service`
33
40
* `/usr/bin/configure-ovs-microshift.sh` for `microshift-ovs-init.service`
34
41
* `/usr/bin/configure-ovs.sh` for `microshift-ovs-init.service`
35
-
* `/etc/crio/crio.conf.d/microshift-ovn.conf` for the CRI-O service
42
+
* `/etc/crio/crio.conf.d/microshift-ovn.conf` for the CRI-O service
Copy file name to clipboardExpand all lines: modules/microshift-nw-topology.adoc
+8-2Lines changed: 8 additions & 2 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -5,7 +5,11 @@
5
5
:_mod-docs-content-type: CONCEPT
6
6
[id="microshift-network-topology_{context}"]
7
7
= Network topology
8
-
OVN-Kubernetes provides an overlay-based networking implementation. This overlay includes an OVS-based implementation of Service and NetworkPolicy. The overlay network uses the Geneve (Generic Network Virtualization Encapsulation) tunnel protocol. The pod maximum transmission unit (MTU) for the Geneve tunnel is set to the default route MTU if it is not configured.
8
+
9
+
[role="_abstract"]
10
+
OVN-Kubernetes provides an overlay-based networking implementation. This overlay includes an OVS-based implementation of `Service` and `NetworkPolicy` resources.
11
+
12
+
The overlay network uses the Geneve (Generic Network Virtualization Encapsulation) tunnel protocol. The pod maximum transmission unit (MTU) for the Geneve tunnel is set to the default route MTU if it is not configured.
9
13
10
14
To configure the MTU, you must set an equal-to or less-than value than the MTU of the physical interface on the host. A less-than value for the MTU makes room for the required information that is added to the tunnel header before it is transmitted.
11
15
@@ -14,13 +18,14 @@ To configure the MTU, you must set an equal-to or less-than value than the MTU o
14
18
The MTU value of the OVN overlay networking in {microshift-short} must be 100 bytes smaller than the MTU value of the base network. If no MTU value is configured, {microshift-short} autoconfigures the value using the MTU value of the default gateway (Internet Protocol version 4 (IPv4) or Internet Protocol version 6 (IPv6)) of the host. If the auto-configuration does not work correctly, the MTU value can be configured manually. For example, if the MTU value of the network is `9000`, the OVN MTU size must be set to `8900`.
15
19
====
16
20
17
-
OVS runs as a systemd service on the {microshift-short} node. The OVS RPM package is installed as a dependency to the `microshift-networking` RPM package. OVS is started immediately when the `microshift-networking` RPM is installed.
21
+
OVS runs as a systemd service on the {microshift-short} node. The OVS RPM package is installed as a dependency to the `microshift-networking` RPM package. OVS starts immediately when the `microshift-networking` RPM is installed.
18
22
19
23
.{product-title} network topology
20
24
image:317_RHbM_OVN_topology_0923.png[title="{microshift-short} uses an overlay-based networking implementation, details follow."]
== Description of the connections in the network topology figure
48
+
43
49
* The north-south traffic between the network service and the OVN external switch `ext_microshift-dev` is provided through the host kernel by the gateway bridge `br-ex`.
44
50
* The OVN gateway router `GR_microshift-dev` is connected to the external network switch `ext_microshift-dev` through the logical router port 4. Port 4 is attached with the node IP address 192.168.122.14.
45
51
* The join switch `join` connects the OVN gateway router `GR_microshift-dev` to the OVN cluster router `ovn_cluster_router`. The IP address range is 100.62.0.0/16.
0 commit comments