Network Services
Core Definitions
| Concept | Description |
|---|---|
| ZCF Network Service Component | The network service module of ZCF. In the UI, it appears as ZNS Network Center and is used to manage ZNS network infrastructure and network service objects. |
| Fabric | The network infrastructure layer of ZNS. It contains objects such as compute managers, hosts, transport zones, uplinks, profiles, and edge clusters. After Fabric is ready, you can continue to configure gateways and policies. |
| Compute Manager | ZNS uses the compute manager to connect to a compute platform and identify compute-side base resources such as hosts and clusters. After a compute manager is registered, ZNS can obtain the host resources required for subsequent network configuration. |
| Host | A compute node that carries service networks and transport node capabilities. After a host is connected to ZNS, you can apply a transport node profile to it, and the host can participate in uplinks, transport zones, and subsequent network configuration. |
| Transport Zone | Defines the host or node scope in which network segments can be transported. Common types include VLAN and Overlay. |
| IP Address Pool | Defines an allocatable IP address range. It is commonly used to provide a TEP address source for host switch profiles and can also serve as the basis for address planning for other network resources. |
| Uplink | A logical uplink object in ZNS. It maps physical NICs on hosts to the Bond-Uplink mapping of a subsequent host switch. |
| Uplink Profile | Defines reusable parameters such as uplink aggregation mode, member count, MTU, Overlay VLAN, and Transport VLAN. |
| Host Switch Profile | Defines host-side switch configuration, including switch type, type, transport zones, IP address pool, uplink profile, and Bond-Uplink mapping. After it is applied to hosts, it affects host-side data-plane network configuration. |
| Transport Node Profile | A host configuration template that references host switch profiles. ZNS administrators can apply this configuration to target hosts. |
| Edge Cluster Profile | A configuration template used to create edge clusters. It contains common edge cluster configuration and description information. |
| Edge Cluster | A set of edge resources that carries network service capabilities such as north-south routing. It can be associated with Tier-0 gateways and Tier-1 gateways. |
| Tier-0 Gateway | A core gateway for external networks or upper-layer routing. It provides edge routing and high availability capabilities. Tier-1 gateways can connect to external networks through a Tier-0 gateway. |
| Tier-1 Gateway | A gateway for tenant or service networks. It can link to a Tier-0 gateway and provide routing capabilities for segments. |
| Segment | A network object in ZNS that carries service network segments. It can be implemented based on a VLAN or Overlay network. After you create a Geneve or VPC network in the ZStack Cloud console, a corresponding segment can be generated in ZNS. |
| NAT Rule | Translates the source or destination IP of traffic on a Tier-0 gateway or Tier-1 gateway. ZNS supports SNAT, DNAT, and Reflexive NAT rules. |
| Load Balancing | Receives TCP or UDP traffic through a VIP on a Tier-1 gateway and forwards the traffic to pool members in a server pool. |
| Virtual Server | The traffic entry point for load balancing. It defines the VIP, listener port, protocol, Tier-1 gateway, server pool, and optional profile. |
| Server Pool | A collection of pool members associated with a Tier-1 gateway. A health monitor can optionally be associated with the server pool. |
| Pool Member | A service endpoint in a server pool that receives traffic. It consists of the IP address of a segment port and a service port. |
| Health Monitor | Checks pool member availability at the configured interval and timeout period, using the configured rise and fall counts. |
| Load Balancing Profile | Defines the persistence criteria and persistence timeout for a virtual server. |
| Router QoS Profile | Used to control bandwidth and burst capability for router-direction traffic. |
| Segment QoS Profile | Used to control ingress and egress traffic rate limiting rules for segments. |
| SpoofGuard Profile | A security profile used to prevent VM IP/MAC address spoofing. |
| SDN Controller | A configuration item in the ZStack Cloud console that is used to connect to ZNS. Its configuration determines how ZStack Cloud networks are associated with ZNS resources such as compute managers, gateways, and segments. |
Prepare ZNS Network Service
To manage network service resources in ZCF, see the ZStack ZNS product documentation to prepare and deploy the ZStack ZNS environment. After the ZStack ZNS environment is prepared, connect it through ZCF Cloud Federation after ZCF is deployed.
Access ZNS Network Center
ZNS administrators can access ZNS Network Center from the Global Management view in ZCF. After access, the left navigation pane displays ZNS management features, including Fabric Overview, Cluster Management, Quick Start, Fabric, IP Management, Connectivity, NAT, Layer 4 Load Balancing, and Network Profiles.
You can view or configure ZNS resources from ZNS Network Center.
View Fabric Overview
The Fabric Overview page shows Fabric resource counts and includes a Resource Summary section, helping ZNS administrators quickly confirm whether the base ZNS resources are ready.
After you confirm that the key resource counts in Fabric Overview match the site plan, you can continue to view cluster status or configure Fabric.
View Cluster Management Status
The Cluster Management page displays the health state of the ZNS control plane cluster and the running state of nodes, VIP, database, OVN, and controllers.
After the cluster, database, OVN, and controllers are healthy, you can continue with Fabric initialization or routine network configuration.
Use Quick Start to Initialize ZNS
The Quick Start page initializes base ZNS resources through a step-by-step wizard. The workflow registers or reuses a compute manager and creates resources such as transport zones, uplinks, IP address pools, profiles, edge clusters, Tier-0 gateways, and Tier-1 gateways.
Note: Quick Start creates or reuses multiple types of ZNS resources. Before you run it, confirm that the compute manager, host NICs, IP address pools, LAG mode, VLAN, and gateway plan are ready.
Note: If a Quick Start task has been started, is running, or has completed, the system automatically returns to the Quick Start status page when you open the Quick Start configuration page. Check the current task progress and step results on the status page first.After Quick Start completes, you can check the generated or reused ZNS resources on the Fabric, IP Management, and Connectivity pages.
Manually Configure Fabric
If the site does not use Quick Start, configure Fabric manually item by item. Manual configuration is suitable when you need fine control over compute managers, host NIC mappings, profiles, edge clusters, and gateway planning.
The recommended configuration order is as follows:
- Register or confirm the compute manager.
- Confirm Fabric global settings and transport zones.
- Create an IP address pool.
- Create uplinks and configure host NIC mappings.
- Create uplink profiles, host switch profiles, and edge cluster profiles.
- Add a transport node profile and apply the configuration to target hosts.
- Create an edge cluster.
Note: Applying configuration to hosts pushes host-side data-plane network configuration. Before you run the operation, confirm that NICs, Bond/LAG, VLAN, IP address pools, and the maintenance window meet site requirements.Register or Confirm a Compute Manager
The Compute Managers page manages registered compute platforms and displays their connection status and configuration state.
After the compute manager is registered, you can view its connection status and configuration state in the list.
Confirm Global Settings and Transport Zones
Before you create host switches, edge clusters, and gateways, confirm Fabric global settings and transport zones. Global settings affect BFD detection and tunnel MTU. Transport zones define the transport scope of VLAN and Overlay segments.
After you confirm that global settings and transport zones match the plan, you can continue to create IP address pools and uplinks.
Create an IP Address Pool
The IP Address Pool page defines network address allocation ranges. When you configure Fabric manually, prepare the IP address pool before you create a host switch profile.
After the IP address pool is created, you can view its name, CIDR, version number, and creation time in the IP address pool list.
Create an Uplink
The Uplinks page records the mapping between logical uplinks and host physical NICs. When you create a host switch profile later, you can select these uplinks in Bond-Uplink Mapping.
After the uplink is created, you can view the uplink name, description, host NIC mappings, and creation time on the Uplinks tab.
Create Profiles
The Profiles page manages edge cluster profiles, host switch profiles, and uplink profiles. Profiles store reusable configurations and reduce repeated input.
After profiles are ready, you can continue to add a transport node profile and apply it to hosts.
Add a Transport Node Profile and Apply It to Hosts
A transport node profile references host switch profiles and is used to apply host-side network configuration to target hosts. Applying configuration affects the host data-plane network. Run the operation only after you confirm the plan.
Note: Applying configuration to hosts pushes related settings such as host switch, uplink, IP address pool, Bond/LAG, and TEP. Run the operation during a maintenance window and confirm the rollback plan in advance.After the operation completes, you can view host connection status, configuration state, and provision state in the host list.
Create an Edge Cluster
The Edge Clusters page manages edge resources that carry routing and network service capabilities. Before you create a Tier-0 gateway, prepare an available edge cluster.
After the edge cluster is created, you can view its name, configuration state, edge cluster profile, host count, description, and creation time in the edge cluster list.
Configure Connectivity
The Connectivity page contains Tier-0 gateways, Tier-1 gateways, and segments. ZNS administrators use Tier-0 and Tier-1 gateways to provide routing capabilities, view service-network objects and deployment state on the Segments page, and configure NAT or Layer 4 Load Balancing on gateways as needed.
When you configure connectivity resources, create or confirm the edge cluster first, then create the Tier-0 gateway, and then create the Tier-1 gateway that links to the Tier-0 gateway. Segments can be generated by the ZStack Cloud network creation workflow. After gateways and service networks are ready, configure NAT rules or Layer 4 load balancing on gateways as needed.
Configure a Tier-0 Gateway
The Tier-0 Gateway page manages core gateways and configures edge routing and high availability mode.
After the Tier-0 gateway is created, you can check whether its configuration state is Deployed in the Tier-0 gateway list.
Configure a Tier-1 Gateway
The Tier-1 Gateway page manages gateways for tenant or service networks and can link to Tier-0 gateways.
After the Tier-1 gateway is created, you can check whether its configuration state is Deployed in the Tier-1 gateway list.
View Segments
The Segments page displays VLAN and Overlay network segments and lets you check configuration state and address information. Some segments can be generated by the ZStack Cloud network creation workflow. ZNS administrators can check deployment results on this page.
After you confirm that the segment is deployed, use the connected gateway, gateway address, IP CIDR, and IP range to troubleshoot service network state.
Network Services
Network services include NAT and load balancing. NAT translates traffic addresses on a Tier-0 gateway or Tier-1 gateway. Load balancing uses a VIP on a Tier-1 gateway to distribute TCP or UDP traffic across multiple pool members.
Before you configure network services, configure the gateways and service networks. For NAT, select a rule type based on the traffic direction. For load balancing, create resources in dependency order: health monitors and profiles, server pools, pool members, and virtual servers.
NAT
NAT rules are associated with a Tier-0 gateway or Tier-1 gateway. Before you create a rule, verify that the target gateway is deployed, select a rule type based on the traffic direction, and prepare the IPv4 addresses, protocol, and ports required for translation.
| Rule Type | Use Case | Translation |
|---|---|---|
| SNAT | Access from an internal network to an external network | Translates the source IP of internal traffic to the source IP used for outbound traffic. |
| DNAT | Access from an external network to an internal service | Translates the destination IP used for external access to the internal service IP and can also translate the destination port. |
| Reflexive NAT | One-to-one bidirectional mapping between an internal IP and an external IP | Creates a bidirectional translation between one internal IPv4 address and one external IPv4 address. |
Create an SNAT Rule
Create an SNAT rule to translate the source IP of matching traffic to an IP used for outbound traffic.
- A Tier-0 gateway or Tier-1 gateway for the SNAT rule is created and deployed.
- The internal IPv4 address or CIDR that requires external access and the translated IPv4 address used for outbound traffic are available.
The SNAT rule appears on the NAT page. Check Address Translation, State, and Deployment Status to verify the configuration.
Create a DNAT Rule
Create a DNAT rule to translate the destination IP or port of external traffic to an internal service IP or port.
- A Tier-0 gateway or Tier-1 gateway for the DNAT rule is created and deployed.
- The IPv4 address for external access and the IPv4 address of the internal service are available. If port translation is required, prepare the external and internal service ports.
The DNAT rule appears on the NAT page. Check Address Translation, State, and Deployment Status to verify the configuration.
Create a Reflexive NAT Rule
Create a Reflexive NAT rule to establish a one-to-one bidirectional translation between an internal IP and an external IP.
- A Tier-0 gateway or Tier-1 gateway for the Reflexive NAT rule is created and deployed.
- The internal and external IPv4 addresses for the one-to-one mapping are available.
The Reflexive NAT rule appears on the NAT page. Address Translation displays the source and translated IPs as a bidirectional mapping.
Manage NAT Rules
As network requirements change, you can edit the translation settings and match criteria of a NAT rule, or enable, disable, or delete the rule. Enable, disable, and delete operations support both individual and bulk actions.
Before you modify a rule, use its Gateway, Rule Type, and Address Translation values to verify that you selected the correct rule and avoid affecting other address translation services.
Edit a NAT Rule
Modify the basic information, apply-to interface, translation settings, or match criteria of a NAT rule.
The NAT rule that you want to modify is created.
The updated rule appears on the NAT page. Check Address Translation, State, and Deployment Status to verify the update.
Enable or Disable NAT Rules
Enable or disable NAT rules to control whether they process NAT traffic.
The NAT rules that you want to enable or disable are created.
State on the NAT page shows the new state. After the rule is deployed, check Deployment Status to verify the operation.
Delete NAT Rules
Delete NAT rules that are no longer required so that they no longer provide address translation.
- Verify that the target rules no longer carry service traffic.
- Use the gateway, rule type, and address translation values to verify the scope of the deletion.
The rules are removed from the NAT page.
Load Balancing
Load balancing provides Layer 4 access to TCP or UDP services through a VIP, protocol, and listener port, and forwards traffic to pool members in a server pool. Load balancing resources are associated with a Tier-1 gateway.
| Resource | Purpose | Requirement |
|---|---|---|
| Health Monitor | Checks whether the service on a pool member is available. | Optional. Create before the server pool. |
| Profile | Maintains sessions based on the source IP, source port, destination IP, or destination port. | Optional. Create before the virtual server. |
| Server Pool | Associates with a Tier-1 gateway and groups pool members that receive traffic. | Required. |
| Pool Member | Consists of the IP address of a segment port and a service port. | Required. Add after the server pool is created. |
| Virtual Server | Defines the VIP, listener port, and protocol that clients use and associates them with a server pool. | Required. Create last. |
For a minimum configuration, you can omit the health monitor and profile. Create a server pool, add pool members, and then create a virtual server.
Create a Health Monitor
To monitor pool member availability, create a health monitor and associate it with a server pool.
The health monitor appears on the Health Monitors tab and is available when you create a server pool.
Create a Load Balancing Profile
To configure session persistence, create a profile that forwards traffic with the same characteristics to the same pool member.
The profile appears on the Profiles tab and is available when you create a virtual server.
Create a Server Pool and Add Pool Members
Create a server pool associated with a Tier-1 gateway and add pool members that receive traffic.
- A Tier-1 gateway for load balancing is created and deployed.
- A service network connected to the Tier-1 gateway is created, and the segment ports of the backend VMs have IPv4 addresses.
The pool members appear on the Pool Members tab for the server pool. Verify each member's IP Address, Port, VM Name, and NIC Name, and verify that State is Enabled.
Create a Virtual Server
Create a virtual server that receives client traffic on a VIP and listener port and forwards the traffic to an associated server pool.
- A server pool is created and contains at least one enabled pool member.
- The VIP, listener port, and protocol that clients use to access the load balancing service are available.
The virtual server appears on the Virtual Servers tab. The list displays VIP, Listener Port, Protocol, Tier-1 Gateway, Server Pool, Profile, and State.
Manage Load Balancing Resources
As service requirements change, you can modify virtual servers, server pools, pool members, health monitors, and profiles. You can also enable or disable virtual servers and pool members to control whether they receive or forward traffic.
Before you detach a segment that contains pool members from a Tier-1 gateway, disable or delete the affected pool members. After the network change, access the VIP from a client to verify connectivity through the load balancing service.
Before you delete a load balancing resource, check its resource relationships and the impact on services. To manage multiple resources, use the bulk actions available on the resource page.
Modify Load Balancing Configuration
Modify the basic information and service settings of load balancing resources.
The load balancing resource that you want to modify is created.
The resource page displays the updated configuration. For a virtual server or pool member, also verify that State meets the service requirements.
Control Load Balancing Traffic
Enable or disable virtual servers and pool members to control load balancing traffic.
The virtual servers or pool members that you want to enable or disable are created.
State on the Virtual Servers or Pool Members page shows the new state.
Prepare Pool Members for Backend Network Changes
Disable or delete affected pool members before you detach their segment from a Tier-1 gateway.
- Identify the segment to change and the affected virtual servers, server pools, and pool members.
- Determine the impact of the network change on load balancing services and schedule an appropriate maintenance window.
Detaching a segment from a Tier-1 gateway does not automatically disable or delete the pool members on that segment. After the segment is detached, Pool Member Status on the virtual server details page might continue to show Up even when clients can no longer access the backend service through the VIP.
Note: Pool Member Status and the pool member State represent different information. After a network topology change, access the VIP from a client to verify that the load balancing service is available. Do not rely solely on Pool Member Status.- Navigate to , and select the Virtual Servers tab.
- Click the target virtual server name. On the virtual server details page, check Pool Member Status and identify the affected pool members by their IP addresses and ports.
- Return to the Load Balancing page, select the Server Pools tab, click the target server pool name, and select the Pool Members tab.
-
Process the affected pool members according to the planned network change.
- For a temporary detachment, select Disable from the Actions column for the target pool member. To process multiple pool members, select them, click Bulk Action, and select Disable.
- For a permanent detachment, select Delete from the Actions column for the target pool member. To process multiple pool members, select them, click Bulk Action, and select Delete.
- In the confirmation dialog box, verify the target pool members and confirm the operation.
- Verify that temporarily disabled pool members show Disabled in the State column, or that permanently deleted pool members are removed from the page.
The affected pool members no longer participate in traffic distribution for the server pool or are removed from the server pool.
- After you process the pool members, follow the applicable network management workflow to detach the segment from the Tier-1 gateway.
- After the network change, access the VIP from a client and verify that the load balancing service meets the service requirements.
Delete Load Balancing Resources
Delete virtual servers, server pools, pool members, health monitors, or profiles that are no longer required.
- Verify that the target resources no longer carry service traffic.
- Check the relationships between the target resources and other load balancing resources.
The resources are removed from the corresponding resource page.
Configure Network Profiles
The Network Profiles page contains Router QoS Profile, Segment QoS Profile, and SpoofGuard Profile. ZNS administrators can create profiles as required by services and select the corresponding profiles when gateways, segments, or service networks use them.
QoS profiles control network traffic bandwidth and burst capability. SpoofGuard profiles prevent VM IP/MAC address spoofing.
Create a Router QoS Profile
The Router QoS Profile page configures traffic shaping rules in the router direction.
After the profile is created, you can view profile name, direction, action, default profile, bandwidth, burst size, description, and creation time in the list.
Note: Default profiles cannot be deleted. Before you delete a profile, confirm that it is no longer used by service gateways.Create a Segment QoS Profile
The Segment QoS Profile page configures ingress and egress rate limiting rules for segments.
After the profile is created, you can view default profile, ingress average bandwidth, ingress peak bandwidth, ingress burst size, egress average bandwidth, egress peak bandwidth, egress burst size, description, and creation time in the list.
Note: Default profiles cannot be deleted. Before you delete a profile, confirm that it is no longer used by service segments.Create a SpoofGuard Profile
The SpoofGuard Profile page prevents VM IP/MAC address spoofing.
After the profile is created, you can view profile name, state, default profile, description, and creation time in the list.
Note: Default profiles cannot be deleted. Before you delete a profile, confirm that it is no longer used by service networks.Relationship with ZStack Cloud Resources
Role Boundary
In ZNS Network Center, ZNS administrators prepare and maintain network infrastructure, gateways, segment state, and network profiles. ZStack Cloud administrators create and use L2 networks, L3 networks, VPC networks, and VMs in the ZStack Cloud console.
When ZStack Cloud networks use ZNS resources, ZNS administrators can check the state of resources such as compute managers, Tier-1 gateways, segments, and network profiles in ZNS Network Center.
Resource Relationship
| Object | What to Check in ZNS | Relationship |
|---|---|---|
| Compute Manager | Confirm that the compute platform is registered, the connection status is Connected, and the configuration state is Deployed. | The ZStack Cloud console connects to ZNS through an SDN Controller. |
| Tier-1 Gateway | Provides uplink and routing capabilities for service networks. Create it according to the service domain or tenant plan. | When you create a ZNS network in the ZStack Cloud console, you can select or associate the target Tier-1 gateway according to the plan. |
| Segment | Check segment name, configuration state, traffic type, VNI/VLAN ID, connected gateway, and address information. | After you create a Geneve or VPC network in the ZStack Cloud console, a corresponding segment might be generated in ZNS. |
| QoS and SpoofGuard Profiles | Create profiles according to service performance and security requirements, and maintain profile state. | After service networks or gateways use profiles, you can check the associated profile state in ZNS. |
Multiple ZStack Cloud Environments
When multiple ZStack Cloud environments connect to the same ZNS environment, ZNS administrators must plan Tier-1 gateways according to service isolation or connectivity requirements. If multiple ZStack Cloud environments share the same Tier-1 gateway, the gateway can carry VM networks in the same service domain or VM networks that require connectivity. If multiple ZStack Cloud environments use different Tier-1 gateways, the gateways can isolate different tenants or service domains.
Note: Whether VM networks in different ZStack Cloud environments can communicate also depends on CIDR conflicts, routing policies, uplink networks, and network configuration in the ZStack Cloud console. When you plan Tier-1 gateways, also confirm service isolation, service connectivity, and uplink network requirements.