Configure Network Services
As an Enterprise user, Orchestrator allows you to configure a number of network services across multiple Edges and Profiles.
- In the SD-WAN service of the Enterprise Portal, select .
- The following screen is displayed:
Figure 1. Network Services 
- You can configure the following network services:
- Non SD-WAN Destinations via Gateway
- Non SD-WAN Destinations via Edge
- API Credentials
- Clusters and Hubs
- Configure Netflow Settings
- DNS Services
- Private Network Names
- Prefix Delegation Tags
- Authentication Services
- TACACS Services
- Edge Services
Note: Configuring Network Services is optional and can be configured in any order.
Configure a Non SD-WAN Destination
The Non SD-WAN Destination (earlier known as Non VeloCloud Site (NVS)) functionality consists of connecting a Arista network to an external Network (for example: Zscaler, Cloud Security Service, Azure, AWS, Partner Datacenter and so on). This is achieved by creating a secure Internet Protocol Security (IPsec) tunnel between a Arista entity and a VPN Gateway at the Network Provider.
- Non SD-WAN Destinations via Gateway- Allows an Gateway to establish an IPsec tunnel directly to a Non SD-WAN Destination. Arista supports the following Non SD-WAN Destination configurations through Gateway:
- AWS VPN Gateway
Note: The AWS VPN Gateway type is introduced in the 4.3.0 release.
- Check Point
- Cisco ASA
- Cisco ISR
- Generic IKEv2 Router (Route Based VPN)
- Microsoft Azure Virtual Hub
- Palo Alto
- SonicWALL
- Zscaler
- Generic IKEv1 Router (Route Based VPN)
- Generic Firewall (Policy Based VPN)
Note: Arista supports both Generic Route-based and Policy-based Non SD-WAN Destination from Gateway.
For information on how to configure Non SD-WAN Destinations via Gateway, see Configure Non SD-WAN Destinations via Gateway
- AWS VPN Gateway
- Non SD-WAN Destinations via Edge- Allows an Edge to establish an IPsec tunnel directly to a Non SD-WAN Destination (AWS and Azure Datacenter). Arista supports the following Non SD-WAN Destination configurations through Edge:
- Generic IKEv1 Router (Route Based VPN)
- Generic IKEv2 Router (Route Based VPN)
- Microsoft Azure Virtual Wan
For information on how to configure Non SD-WAN Destinations via Edge, see Configure Non SD-WAN Destinations via Edge
Non SD-WAN DestinationConfiguration Workflow
- Configure a Non SD-WAN Destination Network Service.
- Associate a Non SD-WAN Destination Network Service to a Profile or Edge.
- Configure Tunnel Parameters: WAN link selection and Per tunnel credentials.
- Configure Business Policy.
VPN Workflow
This is an optional service that allows you to create VPN tunnel configurations to access one or more Non SD-WAN Destinations. Arista provides the configuration required to create the tunnel(s) – including creating IKE IPsec configuration and generating a pre-shared key.
Overview
The following figure shows an overview of the VPN tunnels that can be created between the Arista and a Non SD-WAN Destination.

Configure Non SD-WAN Destinations via Gateway
Arista allows the Enterprise users to define and configure a Non SD-WAN Destination instance to establish a secure IPsec tunnel to a Non SD-WAN Destination through an Gateway.
The Orchestrator selects the nearest Gateway for the Non SD-WAN Destination with its configured IP address, using echolocation service.
You can configure Non SD-WAN Destination via Gateway only at the Profile Level and cannot override at the SD-WAN Edge level.
ECMP
To optimize the utilization of the aggregated bandwidth across the ingress interfaces of non-SD-WAN sites, Arista SD-WAN solution incorporates active-active mode support in its gateways.
This can be achieved by enabling the establishment of multiple IPsec tunnels in active-active mode towards non-SD-WAN sites. This configuration allows load balancing of network traffic across tunnels optimizing the flow of distribution.
- Set up tunnels connecting to non-SD-WAN sites with tunnel mode as Active-Active.
- Choose the preferred load balancing algorithm.
- Configure BGP or Static site subnet routes directing traffic to these sites.
Configure a Non SD-WAN Destination of Type AWS VPN Gateway
This service allows you to create VPN tunnel configurations to access one or more Non SD-WAN Destinations. Arista provides the configuration required to create the tunnel(s) – including creating IKE IPsec configuration and generating a pre-shared key.
Overview
The following figure shows an overview of the VPN tunnels that can be created between Arista and a Non SD-WAN Destination.

Optionally, an IP address can be specified for a Secondary VPN Gateway to form a Secondary VPN Tunnel between an Gateway and the Secondary VPN Gateway. Redundant VPN Tunnels can be specified for any VPN tunnels you create.
Configure a Non SD-WAN Destination of the Type AWS VPN Gateway


| Option | Description |
|---|---|
| General | |
| Name | You can edit the previously entered name for the Non SD-WAN Destination. |
| Type | Displays the type as AWS VPN Gateway. You cannot edit this option. |
| Enable Tunnel(s) | Select the toggle button to initiate the tunnel(s) from the SD-WAN Gateway to the AWS VPN Gateway. |
| Tunnel Mode | Active/Hot-Standby mode supports to set up a maximum of 2 tunnel endpoints or Gateways. |
| Active/Active mode supports to set up a maximum of 4 tunnel endpoints or Gateways. All Active tunnels can send and receive traffic through ECMP. | |
| ECMP Load Sharing Method | Flow Load Based (Default) Flow load based algorithm maps the new flow to the path with least number of flows mapped among the available paths to the destination. |
| Hash Load Based algorithm takes input parameters from 5-tuple (SrcIP, DestIP, SrcPort, DestPort, Protocol). These inputs can be any or all or any subset of this tuple based on user configuration. Flow is mapped to the path based on hash value with selected inputs. | |
| VPN Gateway 1 | Enter a valid IP address. |
| VPN Gateway 2 | Enter a valid IP address. This field is optional. |
| VPN Gateway 3 | Enter a valid IP address. This field is optional. |
| VPN Gateway 4 | Enter a valid IP address. This field is optional. |
| Public IP | Displays the IP address of the Primary VPN Gateway. |
| PSK | The Pre-Shared Key (PSK) is the security key for authentication across the tunnel. The Orchestrator generates a PSK by default. If you want to use your own PSK or password, enter it in the text box. |
| Encryption | Select either AES-128 or AES-256 as the AES algorithm key size to encrypt data. The default value is AES-128. |
| DH Group | Select the Diffie-Hellman (DH) Group algorithm from the drop-down menu. This is used for generating keying material. The DH Group sets the strength of the algorithm in bits. The supported DH Groups are 2, 5, and 14. The default value is 2. |
| PFS | Select the Perfect Forward Secrecy (PFS) level for additional security. The supported PFS levels are deactivated, 2, and 5. The default value is 2. |
| Authentication Algorithm | Select the authentication algorithm for the VPN header. Select one of the supported Secure Hash Algorithm (SHA) functions from the drop-down menu:
The default value is SHA 1. |
| IKE SA Lifetime(min) | Time when Internet Key Exchange (IKE) rekeying is initiated for SD-WAN Edges. The minimum IKE lifetime is 10 minutes and maximum is 1440 minutes. The default value is 1440 minutes. |
| IPsec SA Lifetime(min) | Time when Internet Security Protocol (IPsec) rekeying is initiated for Edges. The minimum IPsec lifetime is 3 minutes and maximum is 480 minutes. The default value is 480 minutes. |
| DPD Type | The Dead Peer Detection (DPD) method is used to detect if the Internet Key Exchange (IKE) peer is alive or dead. If the peer is detected as dead, the device deletes the IPsec and IKE Security Association. Select either Periodic or onDemand from the drop-down menu. The default value is onDemand. |
| DPD Timeout(sec) | Enter the DPD timeout value. The DPD timeout value will be added to the internal DPD timer, as described below. Wait for a response from the DPD message before considering the peer to be dead (Dead Peer Detection).
Prior to the 5.1.0 release, the default value is 20 seconds. For the 5.1.0 release and later, see the list below for the default value.
Note: Prior to the 5.1.0 release, you can deactivate DPD by configuring the DPD timeout timer to 0 seconds. However, for the 5.1.0 release and later, you cannot deactivate DPD by configuring the DPD timeout timer to 0 seconds. The DPD timeout value in seconds will get added onto the default minimum value of 47.5 seconds).
|
| Secondary VPN Gateway | Select the Add button, and then enter the IP address of the Secondary VPN Gateway. Select Save Changes.
The Secondary VPN Gateway is immediately created for this site and provisions a Arista VPN tunnel to this Gateway. |
| Redundant Arista Cloud VPN | Select the check box to add redundant tunnels for each VPN Gateway. Changes made to Encryption, DH Group, or PFS of Primary VPN Gateway also apply to the redundant VPN tunnels, if configured. |
| Local Auth Id | Local authentication ID defines the format and identification of the local gateway. From the drop-down menu, choose from the following types and enter a value:
Note: If you do not specify a value, Default is used as the local authentication ID.
|
| Sample IKE / IPsec | Select to view the information needed to configure the Non SD-WAN Destination Gateway. The Gateway administrator should use this information to configure the Gateway VPN tunnel(s). |
| Location | Select Edit to set the location for the configured Non SD-WAN Destination. The latitude and longitude details are used to determine the best Edge or Gateway to connect to in the network. |
| Site Subnets | Use the toggle button to activate or deactivate the Site Subnets. Select Add to add subnets for the Non SD-WAN Destination. If you do not need subnets for the site, select the subnet and select Delete.
Note:
|
Configure a Non SD-WAN Destination of Type Check Point
The Gateway connects to the Check Point CloudGuard service using IKEv1/IPsec. There are two steps to configure a Check Point: Configuring the Check Point CloudGuard service and configuring the Non SD-WAN Destination of type Check Point. You must perform the first step on the Check Point Infinity Portal and the second step on the Orchestrator.
You must have an active Check Point account and login credentials to access Check Point's Infinity Portal.
Configure the Check Point CloudGuard service
Configure a Non SD-WAN Destination of Type Cisco ASA
Follow the below steps to configure a Non SD-WAN Destination of type Cisco ASA in the Orchestrator.
Configure a Non SD-WAN Destination of Type Cisco ISR
Perform the following steps to configure a Non SD-WAN Destination of type Cisco ISR in the Orchestrator:
Configure a Non SD-WAN Destination of Type Generic IKEv2 Router (Route Based VPN)
Perform the followings steps to configure a Non SD-WAN Destination of type Generic IKEv2 Router (Route Based VPN) in the Orchestrator:
Configure a Non SD-WAN Destination of Type Microsoft Azure Virtual Hub
Follow the below steps to configure a Non SD-WAN Destination of type Microsoft Azure Virtual Hub in the Orchestrator.
- Ensure you have configured a Cloud subscription. For steps, see Configure API Credentials
- Ensure you have created Virtual WAN and Hubs in Azure. For steps, see Configure Azure Virtual WAN for Branch-to-Azure VPN Connectivity
Configure a Non SD-WAN Destination of Type Palo Alto
Follow the below steps to configure a Non SD-WAN Destination of type Palo Alto in the Orchestrator.
Configure a Non SD-WAN Destination of Type Sonic WALL
Perform the below steps to configure a Non SD-WAN Destination of type SonicWALL in the Orchestrator.
Zscaler and VeloCloud SD-WAN Integration
Enterprises can take advantage of secure local Internet breakout by using VeloCloud SD-WAN integrated with Zscaler. Using VeloCloud SD-WAN, the network administrator can decide what traffic should be forwarded to Zscaler, using IPsec tunnels (with NULL encryption).
Prerequisites
- Zscaler Internet Access (ZIA)
- A working instance of ZIA (any cloud)
- Administrator login credentials
- Arista Edge Cloud Orchestrator
- Enterprise account access to Arista Edge Cloud Orchestrator
- Administrator login credentials
- One or more VeloCloud Edge appliances with “Online” status in Arista Edge Cloud Orchestrator
Zscaler Gateway Selection and Routing Behavior

To set the Zscaler tunnel to a specific Gateway, you must first locate which Gateway has the tunnel by following the process above. From there you can select “Secure VPN Gateway” and move/assign the tunnel to a different Gateway.
- Locate current tunnel location.
Figure 26. Tunnel Location 
- Select the Secure VPN Gateway.
Figure 27. Secure VPN Gateway 
- Select a Gateway.
Figure 28. Select Gateway
Note: Assigning/Moving a tunnel to a different Gateway is service affecting. The existing tunnel connection will terminate and a new tunnel from the newly assigned Gateway will be established.During the VeloCloud Edge configuration/activation process, each Edge is assigned a pair of cloud Gateways or a set of Partner Gateways, in accordance with the device configuration. If the Gateways used by the Edge are not the same Gateways which contain the Zscaler tunnels, the Edge will automatically build VCMP tunnels to the Gateways that connect to Zscaler in addition to the Gateways that are selected during the activation process. This ensures the Edge has a path to reach Zscaler.
Zscaler Setup Examples


In this example, only one Zscaler VPN tunnel is created, and the Redundant Velocloud Cloud VPN checkbox is not selected. A single Gateway (Primary Gateway in this case) selected based on the proximity to the remote VPN Gateway (as determined via Geo-IP lookup), will create an IPsec tunnel to the Zscaler VPN endpoint. Dependent on Business Policy configuration, traffic will flow from the Edge, to the Primary Gateway and then on to Zscaler. Even though the Edge always has VCMP tunnels to at least two Gateways, there is no redundancy in this design. Since the Redundant Velocloud Cloud VPN checkbox is not selected, there will not be a backup Gateway tunnel to Zscaler. If either Zscaler or the primary Gateway fails or if the IPsec tunnel between the two goes down for any reason traffic to Zscaler will be dropped.


In this example, only one Zscaler VPN tunnel is created, and the Redundant Velocloud Cloud VPN checkbox is selected. Two Gateways selected based on the proximity to the remote VPN Gateway (as determined via Geo-IP lookup) that are the closest to the Zscaler location will build IPsec tunnels to Zscaler. Both of these tunnels are active, however all traffic to Zscaler will traverse through the Primary Gateway. If the Primary Gateway fails traffic will then shift to the Secondary Gateway. Since only a single Zscaler endpoint is defined if it goes down traffic to Zscaler will be dropped.


In this example, redundant IPsec tunnels to Zscaler are configured in the Orchestrator by adding a secondary Zscaler IP address, however Redundant Velocloud Cloud VPN checkbox is not selected. A single Gateway selected based on the proximity to the remote VPN Gateway (as determined via Geo-IP lookup), will create an IPsec tunnel to both Zscaler VPN endpoints. Both of these tunnels are active, but by configuration settings the Gateway knows which IPsec tunnel to Zscaler is the primary path and will send traffic through that tunnel. Zscaler does not mark primary or backup IPsec tunnels. Zscaler will simply return traffic via the Gateway that originated the request. Should the primary Zscaler location go down, traffic from the Gateway will shift to the secondary Zscaler IPsec tunnel. Since the Redundant Velocloud Cloud VPN checkbox is not selected, there are no redundant Gateway connections to Zscaler. If the Gateway fails, then traffic to Zscaler will be dropped.


In this example, redundant IPsec tunnels to Zscaler are configured in the Orchestrator by adding a secondary Zscaler IP address and Redundant Velocloud Cloud VPN checkbox is selected. Two Gateways selected based on the proximity to the remote VPN Gateway (as determined via Geo-IP lookup), will create IPsec tunnels to both Zscaler VPN endpoints. All of these tunnels are active, but by configuration settings the Gateways knows which of the two is the primary Gateway and which is secondary. The Gateways also know which of their IPsec tunnels to Zscaler is the primary path and which is the secondary path. Zscaler does not mark primary or backup IPsec tunnels. Zscaler will simply return traffic via the Gateway that originated the request. Should the primary Zscaler location go down, traffic from the primary Gateway will shift to the secondary Zscaler IPsec tunnel. Since the Redundant Velocloud Cloud VPN checkbox is selected, if the primary Gateway fails traffic will shift to the secondary Gateway. The secondary Gateway will utilize the primary IPsec tunnel provided that path is available. If not, it will use the secondary IPsec tunnel to reach Zscaler.
Layer 7 Health Checks
When you establish an IPsec/GRE tunnel to a given Zscaler datacenter for Zscaler Internet Access (ZIA), the tunnel is established between the Edge or Gateway, to a virtual IP (VIP) on a Zscaler load balancer for ZIA. When the end user traffic from the branch reaches the load balancer, the load balancer distributes traffic to ZIA Public Service Edges. Dead Peer Detection (DPD) and GRE keepalives can only detect the availability to the public VIP on the load balancer (since it is the tunnel destination). The public VIP is a highly available endpoint and does not reflect the availability of a given ZIA Public Service Edge. Layer 7 health checking allows you to monitor performance and availability of ZIA Edges based on HTTP probes and allows you to failover to an alternate tunnel based on the results. The Edge or Gateway sends probe requests periodically to the HTTP probe URL (in the following format) if probe is activated.
http://gateway.<zscaler_cloud>.net/vpntestThe probe URL is configurable in the Orchestrator, but the probe interval and number of retries are currently not editable in the Orchestrator. If the probe fails consecutively for the number of retries defined, the tunnel is marked down, and the traffic will failover to the secondary tunnel if defined. The probe failure could be either because the https response (200 OK) is not received, or the latency is greater than the defined threshold. If conditional backhaul is configured in an Edge, probe failures to both primary and secondary tunnel will trigger traffic failover to the backhaul hub configured. When the probe is UP again, traffic will fall back to the CSS tunnel. If Redundant Cloud VPN is configured for Non SD-WAN Destination (NSD) via Gateway, probe failures to both primary and secondary tunnel from primary gateway will trigger traffic failover to secondary gateway. When the probe in the primary gateway is UP again, traffic will fall back to the CSS tunnel on the primary gateway.
Zscaler and VeloCloud SD-WAN Deployment Configurations
Layer 7 health check Events
| Event | Displayed on Orchestrator UI as | Severity | Notification Configurable | Generated By | Generated When |
|---|---|---|---|---|---|
| EDGE_NVS_TUNNEL_UP | Edge Direct IPsec tunnel up | INFO | N | Orchestrator | A Cloud Security Service tunnel or NSD via Edge tunnel is up. |
| EDGE_NVS_TUNNEL_DOWN | Edge Direct IPsec tunnel down | INFO | N | Orchestrator | A Cloud Security Service tunnel or NSD via Edge tunnel is down. |
| VPN_DATACENTER_STATUS | VPN Tunnel state change | NOTICE | N | Gateway | The VPN Tunnel state is changed. |
Configure a Non SD-WAN Destination of Type Zscaler
To create and configure a Non SD-WAN Destination of type Zscaler, perform the following steps:
Associate a Non SD-WAN Destination to a Configuration Profile
After configuring a Non SD-WAN Destination of type Zscaler in Orchestrator, you have to associate the Non SD-WAN Destination to the desired Profile in order to establish the tunnels between Gateways and Zscaler VPN Gateways. To associate a Non SD-WAN Destination to a configuration profile, perform the following steps:
Configure Zscaler
This section discusses the Zscaler configuration.
Perform the below steps on the Zscaler website. From there, create a Zscaler account, add VPN credentials, and add a location.
Configure Business Priority Rules
Define the business policy in your Orchestrator to determine web security screening. The business policy matches parameters such as IP addresses, ports, VLAN IDs, interfaces, domain names, protocols, operating system, object groups, applications, and DSCP tags. When a data packet matches the match conditions, the associated action or actions are taken. If a packet matches no parameters, then a default action is taken on the packet.
You can configure Business Policy rules using the Business Policy tab in the Profile Configuration page. Optionally, you can also override the Profile Business Policy rules at the Edge-level. To create a business policy at the Edge level:
Configure a Non SD-WAN Destination of Type Generic IKEv1 Router (Route Based VPN)
To configure a Non SD-WAN Destination of type Generic IKEv1 Router (Route Based VPN) in the Orchestrator, perform the below steps:.
Configure a Non SD-WAN Destination of Type Generic Firewall (Policy Based VPN)
Follow the below steps to configure a Non SD-WAN Destination of type Generic Firewall (Policy Based VPN) in the Orchestrator.
Configure Non SD-WAN Destinations via Edge
Arista allows the Enterprise users to define and configure a Non SD-WAN Destination instance in order to establish a secure IPSec v4 and v6 tunnels directly from an Edge to a Non SD-WAN Destination. This section also allows you to configure Cloud Security Services.
- Configure tunnel settings for your Non SD-WAN Destination. For additional information, see:
- Associate your Non SD-WAN Destination to a Profile or Edge. For additional information, see Configure a Tunnel Between a Branch and a Non SD-WAN Destinations via Edge.
- Configure Tunnel parameters (WAN link selection and Per tunnel credentials) at the Edge level. For additional information, see Configure Cloud VPN and Tunnel Parameters for Edges.
- Configure Business Policy. Configuring business policy is an optional procedure for Non SD-WAN Destinations via Edge. If there are no Non SD-WAN Destinations configured then you can redirect the Internet traffic via business policy. For additional information, see Create Business Policy Rule.
Configure a Non SD-WAN Site of Type Generic IKEv1 Router via Edge
This topic discusses how to configure a Non SD-WAN Destination of type Generic IKEv1 Router (Route Based VPN) through Edge in Orchestrator.
Configure a Non SD-WAN Site of Type Generic IKEv2 Router via Edge
This topic discusses how to configure a Non SD-WAN Destination of type Generic IKEv2 Router (Route Based VPN) through Edge in Orchestrator.
Configure a Non SD-WAN Site of Type Microsoft Azure via Edge
- Ensure you have configured a Cloud subscription. For steps, see Configure API Credentials.
- Ensure you have created Virtual WAN and Hubs in Azure. For steps, see Configure Azure Virtual WAN for Branch-to-Azure VPN Connectivity.
To configure a Non SD-WAN Destination of type Microsoft Azure Virtual Hub via Edge in Orchestrator:
- Configure Cloud VPN for Profiles
- Associate the Microsoft Azure Non SD-WAN Destination to an Edge and configure tunnels to establish a tunnel between a branch and Azure Virtual Hub. For additional information, see Associate a Microsoft Azure Non SD-WAN Destination to an Edge and Add Tunnels.
For information about Azure Virtual WAN Edge Automation, see Configure Orchestrator for Azure Virtual WAN IPsec Automation from Edge.
Configure Tunnel Between Branch and Non SD-WAN Destinations via Edge
After configuring a Non SD-WAN Destination via Edge in Orchestrator, you have to associate the Non SD-WAN Destination to the desired Profile in order to establish the tunnels between Gateways and the Non SD-WAN Destination.
To establish a VPN connection between a branch and a Non SD-WAN Destination configured via Edge, perform the following steps:
Configure API Credentials
This section allows you to configure both, Iaas and Cloud Subscriptions. Iaas Subscription refers to Microsoft Azure Subscription and Cloud Subscription refers to Zscaler Subscription.
Configure Clusters and Hubs
This section allows you to configure Edge Clusters. You can also view the existing Cloud VPN Hubs.
About Edge Clustering
The size of a single VPN Network with a VeloCloud Hub is constrained by the scale of the individual Hub. For large networks containing thousands of remote sites, it would be preferable for both scalability and risk mitigation to use multiple Hubs to handle the Edges. However, it is impractical to mandate that the customer manage individual separate Hubs to achieve this. Clustering allows multiple Hubs to be leveraged while providing the simplicity of managing those Hubs as one common entity with built-in resiliency.
Edge Clustering addresses the issue of Hub scale because it can be used to easily expand the tunnel capacity of the Hub dynamically by creating a logical cluster of Edges. Edge Clustering also provides resiliency via the Active/Active High Availability (HA) topology that a cluster of Edge would provide. A cluster is functionally treated as an individual Hub from the perspective of other Edges.
The Hubs in a Cluster can be either physical or Virtual Edges. If they are virtual, they may exist on a single hypervisor or across multiple hypervisors.
Each Edge in a cluster periodically reports usage and load stats to the Gateway. The load value is calculated based on Edge CPU and memory utilization along with the number of tunnels connected to the Hub as a percentage of the Edge model’s tunnel capacity. The Hubs within the cluster do not directly communicate nor exchange state information. Typically, Edge Clusters are deployed as Hubs in data centers.
How Edge Clustering Works
This section provides an in-depth overview of how the SD-WAN Edge Clustering functionality works.
- Edge Clustering can be used on Hubs as follows:
- To allow greater tunnel capacity for a Hub than an individual Edge serving as a Hub can provide.
- To distribute the remote Spoke Edges among multiple Hubs and reduce the impact of any incident that may occur.
- Cluster Score is a mathematical calculation of the overall utilization of the system as follows:
The three measured utilization factors are CPU usage, memory usage, and tunnel capacity.
- Each measure of utilization is treated as a percentage out of a maximum of 100%.
- Tunnel capacity is based on the rated capacity for a given hardware model or Virtual Edge configuration.
- All three utilization percentages are averaged to arrive at an integer-based Cluster Score (1-100).
- While throughput is not directly considered, CPU and memory usage indirectly reflect throughput and flow volume on a given Hub.
- For example, on an Edge 2000:
- CPU usage = 20%
- Memory usage = 30%
- Connected Tunnels = 600 (out of a capacity of 6000) = 10%
- Cluster Score: (20 + 30 + 10)/3 = 20
- A Cluster Score greater than 70 is considered "over capacity."
- A “logical ID” is a 128-bit UUID that uniquely identifies an element inside the Arista Network.
- For instance, each Edge is represented by a logical ID and each Cluster is represented by a logical ID.
- While the user is providing the Edge and Cluster names, the logical IDs are guaranteed to be unique and are used for internal identification of elements.
- By default, the load is evenly distributed among Hubs. Hence, it is necessary that all Edges that are part of a cluster must be of the same model and capacity.
The Hub Cluster members must learn the exact Spoke prefixes from the other Cluster members. This is required for Dynamic Branch to Branch via Hub tunnels to be established between Spokes when the Spokes are connected to different Cluster members. The source Spoke checks the reachability of the destination Spoke's exact route via the source Spoke's Hub, so if the Hub doesn't have this exact route, then it is marked as unreachable, and the traffic fails.
How are Edge Clusters Tracked by the VeloCloud Gateway?
Once a Hub is added to a VeloCloud SD-WAN Cluster, the Hub will tear down and rebuild tunnels to all of its assigned Gateways and indicate to each Gateway that the Hub has been assigned to a Cluster and provide a Cluster logical ID.

- The logical ID
- The name
- Whether Auto Rebalance is activated
- A list of Hub objects for members of the Cluster
- The logical ID
- The name
- A set of statistics, updated every 30 seconds via a periodic message sent from the Hub to each assigned Gateway, including:
- Current CPU usage of the Hub
- Current memory usage of the Hub
- Current tunnel count on the Hub
- Current BGP route count on the Hub
- The current computed Cluster Score based on the formula provided above.
A Hub is removed from the list of Hub objects when the Gateway has not received any packets from the Hub Edge for more than seven seconds.
How are Edges Assigned to a Specific Hub in a Cluster?
In a traditional Hub and Spoke topology, the Orchestrator provides the Edge with the logical ID of the Hub to which it must be connected. The Edge asks its assigned Gateways for connectivity information for that Hub logical ID—i.e. IP addresses and ports, which the Edge will use to connect to that Hub.
From the Edge’s perspective, this behavior is identical when connecting to a Cluster. The Orchestrator informs the Edge that the logical ID of the Hub it should connect to is the Cluster logical ID rather than the individual Hub logical ID. The Edge follows the same procedure of sending a Hub connection request to the Gateways and expects connectivity information in response.

- Divergence Number One: The Gateway must choose which Hub to assign.
- Divergence Number Two: Due to Divergence Number One, the Edge may get different assignments from its different Gateways.
Divergence Number One was originally addressed by using the Cluster Score to assign the least loaded Hub in a Cluster to an Edge. While in practice this is logical, in the real world, it turned out to be a less than ideal solution because a typical reassignment event can involve hundreds or even thousands of Edges and the Cluster Score is only updated every 30 seconds. In other words, if Hub 1 has a Cluster Score of 20 and Hub 2 has a Cluster Score of 21, for 30 seconds all Edges would choose Hub 1, at which point it may be overloaded and trigger further reassignments.
- Edge logical ID modulo the number of Hubs in Cluster = Assigned Hub index
- For example:
- Four Edges that have logical IDs ending in 1, 2, 3, 4
- Cluster with 2 Hubs
- 1 % 2 = 1, 2 % 2 = 0, 3 % 2 = 1, 4 % 2 = 0 (Note: "%” is used to indicate the modulo operator)
- Edges 2 and 4 are assigned Hub Index 0
- Edges 1 and 3 are assigned Hub Index 1
-
Figure 67. Cluster with 2 Hubs 
This is more consistent than a round-robin type assignment because it means that Edges will tend to be assigned the same Hub each time, which makes assignment and troubleshooting more predictive.
What Happens when a Hub Exceeds its Maximum Allowed Tunnel Capacity?
The Edge assignment logic will attempt to evenly distribute the Edges between all available Hubs. However, after an event (like restart) on the Hub, the Edge distribution will no longer be even.
Due to such an event on the Hub (or adding additional Edges to the network), Clusters might reach a point where an individual Hub has exceeded 70% of its permitted tunnel capacity. If this happens, and at least one other Hub is at less than 70% tunnel capacity, then fair share redistribution is performed automatically regardless of whether rebalancing is activated on the Orchestrator. Most Edges will retain their existing assignment due to the predictive mathematical assignment using logical IDs, and the Edges that have been assigned to other Hubs due to failovers or previous utilization rebalancing will be rebalanced to ensure the Cluster is returned to an even distribution automatically.
What Happens when a Hub Exceeds its Maximum Allowed Cluster Score?
Unlike tunnel percentage (a direct measure of capacity), which can be acted upon immediately, the Cluster Score is only updated every 30 seconds and the Gateway cannot automatically calculate what the adjusted Cluster Score will be after making an Edge reassignment. In the Cluster configuration, an Auto Rebalance parameter is provided to indicate whether the Gateway should dynamically attempt to shift the Edge load for each Hub as needed.
If Auto Rebalance is deactivated and a Hub exceeds a 70 Cluster Score (but not 70% tunnel capacity), then no action is taken.
If Auto Rebalance is activated and one or more Hubs exceed a 70 Cluster Score, the Gateway will reassign one Edge per minute to the Hub with the lowest current Cluster Score until all Hubs are below 70 or there are no more reassignments possible.
What Happens when two VeloCloud Gateways give Different Hub Assignments?
As is the nature of a distributed control plane, each Gateway is making an individual determination of the Cluster assignment. In most cases, Gateways will use the same mathematical formula and thus arrive at the same assignment for all Edges. However, in cases like Cluster Score-based rebalancing this cannot be assured.
If an Edge is not currently connected to a Hub in a Cluster, it will accept the assignment from any Gateway that responds. This ensures that Edges are never left unassigned in a scenario where some Gateways are down and others are up.
What Happens when a VeloCloud Gateway Goes Down?
When a SD-WAN Gateway goes down, Edges may be reassigned if the most preferred Gateway was the one that went down, and the next most preferred Gateway provided a different assignment. For instance, the Super Gateway assigned Hub A to this Edge while the Alternate Super Gateway assigned Hub B to the same Edge.
The Super Gateway going down will trigger the Edge to fail over to Hub B, since the Alternate Super Gateway is now the most preferred Gateway for connectivity information.
When the Super Gateway recovers, the Edge will again request a Hub assignment from this Gateway. In order to prevent the Edge switching back to Hub A again in the scenario above, the Hub assignment request includes the currently assigned Hub (if there is one). When the Gateway processes the assignment request, if the Edge is currently assigned a Hub in the Cluster and that Hub has a Cluster Score less than 70, the Gateway updates its local assignment to match the existing assignment without going through its assignment logic. This ensures that the Super Gateway, on recovery, will assign the currently connected Hub and prevent a gratuitous failover for its assigned Edges.
What Happens if a Hub in a Cluster Loses its Dynamic Routes?
As noted above, the Hubs report to the SD-WAN Gateways the number of dynamic routes they have learned via BGP every 30 seconds. If routes are lost for only one Hub in a Cluster, either because they are erroneously retracted or the BGP neighborship fails, the SD-WAN Gateways will failover Spoke Edges to another Hub in the Cluster that has an intact routing table.
As the updates are sent every 30 seconds, the route count is based on the moment in time when the update is sent to the SD-WAN Gateway. The SD-WAN Gateway rebalancing logic occurs every 60 seconds, meaning that users can expect failover to take 30-60 seconds in the unlikely event of total loss of a LAN-side BGP neighbor. To ensure that all Hubs have a chance to update the Gateways again following such an event, rebalancing is limited to a maximum of once per 120 seconds. This means that users can expect failover to take 120 seconds for a second successive failure.
- Routes received from BGP over IPsec/GRE are not accounted for LAN side failure detection. When BGP over IPsec/GRE session goes down, the issue is not detected by LAN side failure and therefore this does not trigger cluster failover.
- BGP routes learned from a BGP neighbor connected to a routed interface with the "Enable WAN Link" option enabled, are not counted towards the number of dynamic routes the Hub reports to the SD-WAN Gateways. Only LAN-side BGP routes are reported and "Enable WAN Link" implies that the link is on the WAN-side.
How to Configure Routing on Cluster Hubs?
As the Gateway can instruct the spokes to connect to any member Hub of the Cluster, the routing configuration should be mirrored on all the Hubs. For example, if the spokes must reach a BGP prefix 192.168.2.1 behind the Hubs, all the Hubs in the cluster should advertise 192.168.2.1 with the exact same route attributes.
BGP uplink community tags should be used in the cluster deployment. Configure the cluster nodes to set the uplink community tag when redistributing routes to BGP peers.
What Happens if a Hub in a Cluster Fails?
The SD-WAN Gateway will wait for tunnels to be declared dead (7 seconds) before failing over Spoke Edges. This means that users can expect failover to take 7-10 seconds (depending on RTT) when an SD-WAN Hub or all its associated WAN links fail.
Troubleshooting Edge Clustering
This section discusses the troubleshooting enhancements for Edge Clustering.
Overview
Edge Clustering includes a troubleshooting feature to rebalance VeloCloud SD-WAN Spoke Edges within a Cluster. The rebalancing of the Spokes can be performed on any of the Hubs within the Cluster. There are two methods to rebalance Spokes:
- Evenly rebalance Spokes across all the Hubs in the Cluster.
- Exclude one Hub and rebalance the Spokes across the remaining Hubs in the Cluster.
Rebalancing Spokes on the Hub Using the VeloCloud Orchestrator
An administrator may rebalance Spokes in a Cluster via Remote Diagnostics on the VeloCloud Orchestrator. When an Edge is deployed as a Hub in a Cluster, a new Remote Diagnostics option will appear named Rebalance Hub Cluster, which offers users two choices.
Redistribute Spokes in Hub Cluster
- This option attempts to evenly re-distribute Spoke Edges among all Hub Edges in the Cluster.
Redistribute Spokes Excluding this Hub
- This option attempts to evenly re-distribute Spokes among Hubs in the Cluster, excluding the Hub Edge from which a user is running the Redistribute Spokes utility.
- This option can be used for troubleshooting or maintenance to remove all Spokes from this Hub Edge.
Hub or Cluster Interconnect
- Ensure to upgrade the Orchestrator, Gateways, and Hubs or Hub Clusters to version 5.4.0.0 or above.
- The Cloud VPN service must be activated for the Cluster Profile associated with the Edge Clusters or Hubs.
- The Branch to Branch VPN (Transit & Dynamic) check box must not be selected in interconnect Hub Profiles, as shown below.
Figure 68. Disabling the Branch to Branch VPN 
Configuring Hubs Designation on interconnect Profiles is sufficient for end to end communication with all nodes. You can configure the Branch to Branch via Hubs for Spoke Profiles.
- The Hub or Cluster Interconnect feature must be activated in all the Hub Profiles involved in the interconnect process.
- Cluster members must run the BGP with LAN/L3 router, and the router must be configured to forward the BGP extended communities.
- There must be at least one common Gateway for all Edges (Spokes and Hubs) in case of Partner Gateway assignment. The order of Partner Gateways assignment should be same across all the Hub/Cluster Profiles.
VeloCloud supports interconnection of multiple Hub Edges and/or Hub Clusters to increase the range of Spoke Edges that can communicate with each other. This feature allows communication between the Spoke Edges connected to one Hub Edge/Hub Cluster and the Spoke Edges connected to another Hub Edge/Hub Cluster, using multiple overlay and underlay connections.
When a Spoke Edge tries to connect to a Hub Cluster, one of the members from the Hub Cluster is selected as the Hub to the Spoke Edge. If this Hub goes down, another member from the same Hub Cluster is automatically selected to serve the Spoke Edge, without any user configuration. The Hub Cluster members are connected to each other via underlay (BGP), and can exchange the routes and data using this underlay connection. Spoke Edges connected to different members of the same Hub Cluster can then communicate with each other using this underlay connection. This solution provides better resiliency.

- The Hub or Cluster Interconnect feature must be activated.
- The Branch to Hub Site (Permanent VPN) check box must be selected. The two interconnected Hub nodes must be configured as Hubs to each other as explained in the below table.
| Profile | Hubs Designation |
|---|---|
| hub_profile1 | hub2 |
| hub_profile2 | hub1 and hub3 |
| hub_profile3 | hub2 |
When the Hub or Cluster Interconnect feature is activated, tunnels are formed from one Cluster to another Cluster with at least one peer in other Cluster. Based on the condition, two members from one Cluster can form tunnels to same members in another Cluster. In case of individual Hub and Hub Cluster interconnect, all the Cluster members form tunnels to that individual Hub. The end Spoke Edges connected to these Hub Clusters can then communicate with each other through these two Hub Clusters and the intermediate VeloCloud SD-WAN Routing Protocol hops.
The intra Cluster routes are advertised with special BGP extended community, wherein the last four bytes of the Cluster ID are embedded in the extended community. For example, if the Cluster ID is fee2f589-eab6-4738-88f2-8af84b1a3d9c, 4b1a3d9c is reversed and used to derive the Cluster community as 9c3d1a4b00000003. Based on this community tag, the intra Cluster routes are filtered out towards the controller. This avoids reflecting redundant routes from multiple Cluster members.

- Overlay connection between S1 and C1.
- Overlay connection between S2 and C2.
- Overlay connection between C1 and C2.
- Underlay connection within C1.
- Underlay connection within C2.
In this way, the Hub Clusters can exchange routes with each other, providing a way for the packets to flow between Spoke Edges connected to different Hub Clusters.
- Dynamic branch to branch is supported between Spokes connected to two different or same Clusters.
- Profile isolation in Spoke Profile is supported.
- Internet Backhaul via Cluster is supported.
Limitations:
- Hub or Cluster Interconnect through Gateway is not supported.
- Exchanging routes between Hub Cluster members using OSPF is not supported.
- Asymmetric routing can occur when two Clusters are interconnected. Enhanced Firewall Services or Stateful Firewall must not be activated as they can block the traffic due to asymmetric routing.
- When all the Overlay tunnels go down between two Cluster members, traffic drop is expected until they form a tunnel with other members in the peer Cluster.
- If there are more than one LAN/WAN routers running BGP with Cluster, Trusted Source check box must be selected and the value of Reverse Path Forwarding must be Not enabled, on the Cluster Edge interfaces connecting BGP routers. For more information, see Configure Interface Settings for Edges.
- Without the Hub or Cluster Interconnect feature, a Cluster Profile cannot have another Cluster or Hub configured as a Hub.
Configuring Hub or Cluster Interconnect
Configure Netflow Settings
In an Enterprise network, Netflow monitors traffic flowing through Edge and exports Internet Protocol Flow Information Export (IPFIX) information directly from Edge to one or more Netflow collectors. IPFIX is an IETF protocol that defines the standard of exporting flow information from an end device to a monitoring system. Arista VeloCloud supports IPFIX version 10 to export IP flow information to a collector. Generally, an IP flow is identified by five tuples namely: Source IP, Destination IP, Source Port, Destination Port, and Protocol. But the Netflow records that are exported by Edge aggregates the source port. This means that data of different flows that have same source and destination IPs, same destination port, but different source ports will be aggregated.
The Orchestrator allows you to configure Netflow collectors and filters as network services at the Profile, Edge, and Segment level. You can configure a maximum of two collectors per Segment and eight collectors per Profile and Edge. Also, you can configure a maximum of 16 filters per collector.
- While configuring a Profile or Edge, you can either select a collector and filter from the available list or add a new collector and a filter. For steps, see Configure Netflow Settings for Profiles.
- To override Netflow settings at the Edge level, see Configure Netflow Settings for Edges.
After you enable Netflow on the Edge, it periodically sends messages to the configured collector. The contents of these messages are defined using IPFIX templates. For additional information on templates, see IPFIX Templates.
IPFIX Templates
After you enable Netflow on the VeloCloud Edge, it periodically sends messages to the configured collector. The contents of these messages are defined using templates. Internet Protocol Flow Information Export (IPFIX) templates have additional parameters that provide more information regarding the traffic flows.
Non-NAT Template
sourceIPv4AddresdestinationIPv4AddressdestinationTransportPortingressVRFIDApplicationIDprotocolIdentifier
Source port is aggregated out.
The Non-NAT template is the common Netflow template.
| Element ID | Name | Type | Description | Recommended Implementation | Applicable Edge Release |
|---|---|---|---|---|---|
| 1 | octetDeltaCount | unsigned64 | The number of octets includes IP header(s) and IP payload. | Used to report on total bytes (aggregate of bytesTX and bytesRx) and BytesRX. | 3.3.0 |
| 2 | packetDeltaCount | unsigned64 | The number of incoming packets since the previous report (if any) for this flow at the observation point. | Used to report on total packet (aggregate of packetTX and packetRX) and packetRX. | 3.3.0 |
| 32769 | octetDeltaCount_rev | unsigned64 | Biflow RFC5103. The number of outgoing byte. | Used to report on total bytes (aggregate of bytesTX and bytesRX) and BytesTX. | 3.3.0 |
| 32770 | packetDeltaCount_rev | unsigned64 | Biflow RFC5103. The number of outgoing packets. | Used to report on total packet (aggregate of packetTX and packetRx) and packetTX. | 3.3.0 |
| 3 | deltaFlowCount | unsigned64 | The conservative count of original flows contributing to this aggregated flow; may be distributed via any of the methods expressed by the valueDistributionMethod Information Element. | See IPFIX Information Element Definitions. | 3.3.0 |
| 4 | protocolIdentifier | unsigned8 | The value of the protocol number in the IP packet header. The protocol number identifies the IP packet payload type. Protocol numbers are defined in the IANA Protocol Numbers registry. | Implement as per description. | 3.3.0 |
| 5 | ipClassOfService | unsigned8 | For IPv4 packets, this is the value of the TOS field in the IPv4 packet header. | Implement as per description. | 3.3.0 |
| 8 | sourceIPv4Address | ipv4Address | The IPv4 source address in the IP packet header. | Implement as per description. | 3.3.0 |
| 10 | ingressInterface | unsigned32 | The index of the IP interface where packets of this flow are being received. The value matches the value of managed object 'ifIndex' as defined in RFC2863. | This value maps to Interface option template 272 ‘ingressInterface’ value where to map the flow to SD-WAN link interface number. | 3.3.0 |
| 11 | destinationTransportPort | unsigned16 | The destination port identifier in the transport header. | Implement as per description. | 3.3.0 |
| 12 | destinationIPv4Address | ipv4Address | The IPv4 destination address in the IP packet header. | Implement as per description. | 3.3.0 |
| 14 | egressInterface | unsigned32 | The index of the IP interface where packets of this flow are being sent. The value matches the value of managed object 'ifIndex' as defined in RFC2863. | Egress interface | 3.3.0 |
| 15 | ipNextHopIPv4Address | ipv4Address | The IPv4 address of the next IPv4 hop. RFC2863 | This IP address identifies the next hop device when there is no SD-WAN overlay (underlay next hop). | 3.3.0 |
| 56 | sourceMacAddress | macAddress | The IEEE 802 source MAC address field. | Implement as per description. | 3.3.0 |
| 239 | biflowDirection | unsigned8 | A description of the direction assignment method used to assign the biflow Source and destination. This Information element may be present in a flow data record or applied to all flows exported from an exporting process or observation domain using IPFIX options. If this Information element is not present in a flow record or associated with a biflow via scope, it is assumed that the configuration of the direction assignment method is done out-of-band.
Note: When using IPFIX options to apply this Information element to all flows within an observation domain or from an exporting process, the option should be sent reliably. If reliable transport is not available (i.e., when using UDP), this Information element should appear in each flow record.
|
See IPFIX Information Element Definitions. | 3.3.0 |
| 95 | applicationId | octetArray(8) | Specifies an application ID. RFC6759. For details, see Application Option Template. | Implement to recognize L7 app signature. | 3.3.0 |
| 148 | flowID | unsigned64 | An identifier of a flow that is unique within an observation domain. This Information element can be used to distinguish between different flows if flow keys such as IP addresses and port numbers are not reported or are reported in separate records. | Unique flow ID maps to flow links stats option template 257. | 3.3.0 |
| 152 | flowStartMilliseconds | dateTime Milliseconds | The absolute timestamp of the first packet of this flow. | Implement as per description. | 3.3.0 |
| 153 | flowEndMilliseconds | dateTime Milliseconds | The absolute timestamp of the last packet of this flow. | Implement as per description. | 3.3.0 |
| 136 | flowEndReason | unsigned8 | The reason for flow termination. The range of values includes the following:
|
Implement as per description. | 3.3.0 |
| 234 | ingression | unsigned32 | A unique identifier of the Rename where the packets of this flow are being received. This identifier is unique per metering process. | This maps to the VeloCloud Orchestrator segments. A segment should be visualized and reported as a separated L3 domain within the Edge | 3.3.0 |
Enterprise-Specific Fields (ID>32767)
| Element ID (Enterprise Element ID) | Name | Type | Description | Recommended Implementation | Applicable Edge Release |
|---|---|---|---|---|---|
| 45001 (12233) | destinationUUID | octetArray | Destination node UUID | This identifies the final SD-WAN endpoint in the path (same as nexthop UUID in e2e). | 3.3.0 |
| 45002 (12234) | vcPriority | unsigned8 |
|
This identifies the BizPolicy ‘Priority’ classification applied.
Unset should be monitored to deduce a warning since it would only occur during overflow. |
3.3.0 |
| 45003 (12235) | vcRouteType | unsigned8 |
|
This identifies the path type out to Internet the flow is taking.
Unset should be monitored to deduce a warning since it would only occur during overflow. |
3.3.0 |
| 45004 (12236) | vcLinkPolicy | unsigned8 |
|
This value provides the type of link steering and remediation configured for this application under BizPolicy. | 3.3.0 |
| 45005 (12237) | vcTrafficType | unsigned8 |
|
This identifies the BizPolicy ‘Service Class' classification applied. | 3.3.0 |
| 45007 (12239) | vcFlowPath | unsigned8 |
|
This identifies the type of ‘path’ the flow is taking. | 3.3.0 |
| 45009 (12241) | replicatedPackets RxDeltaCount | unsigned64 | Count of replicated packets received for the flow | This value provides the number of packets replicated (FEC) in the Rx path due to loss (applies to real-time protocols). | 3.3.0 |
| 45010 (12242) | replicatedPackets TxDeltaCount | unsigned64 | Count of packets replicated for the flow | This value provides the number of packets replicated (FEC) in the Tx path due to loss (applies to real-time protocols). | 3.3.0 |
| 45011 (12243) | lostPacketsRx DeltaCount | unsigned64 | Count of packets lost for the flow at the receive | This value provides the total number of packets lost for the flow. | 3.3.0 |
| 45012 (12244) | retransmittedPackets TxDeltaCount | unsigned64 | Count of packets retransmitted for the flow | This value provides the number of retransmitted packets due to loss (applies to transactional traffic). | 3.3.0 |
| 45085 (12317) | tcpRttMs | unsigned16 | Maximum RTT observed for a TCP flow | The maximum Roundtrip Time observed in milliseconds for the tcp packets in the flow, since the previous report (if any) for this flow at the observation point. | 4.0.0 |
| 45086 (12318) | tcpRetransmits | unsigned32 | Count of TCP packets retransmitted for the flow | The number of TCP packets retransmitted since the previous report (if any) for this flow at the observation point. | 4.0.0 |
| 45080 (12312) | bizPolicyId | string | Business policy logical Id this flow is matching. | This value is a UUID and must be mapped to a BizPolicy via Orchestrator API. | 3.3.2 |
| 45082 (12314) | nextHopUUID | octetArray | Next hop UUID for this flow. This will be populated in case of overlay traffic. | This value identifies the device that is in the path between source and destination in the SD-WAN overlay network (not underlay). | 3.3.2 |
NAT Template
Common + NAT template
| Element ID | Name | Type | Description | Applicable Edge Release |
|---|---|---|---|---|
| 225 | postNATSourceIPv4Address | ipv4Address | The definition of this information element is identical to the definition of information element sourceIPv4Address, except that it reports a modified value caused by a NAT middlebox function after the packet passed the observation point. | 3.4.0 |
| 226 | postNATDestinationIPv4Address | ipv4Address | The definition of this information element is identical to the definition of information element destinationIPv4Address, except that it reports a modified value caused by a NAT middlebox function after the packet passed the observation point. | 3.4.0 |
- Netflow exports are unidirectional flows. VeloCloud SD-WAN needs to export flow stats as two flow records or implement RFC5103 (Bidirectional Flow Export).
- flowID will need to be constructed to be unique within the Enterprise.
- Direct NAT:
- Consider a flow which comes from LAN client with IP
10.0.1.25to Internet169.254.6.18. This gets NATed due to business policy (SNAT source IP to a WAN interface IP169.254.7.10). So, flow record for this flow will be with SIP:10.0.1.25and DIP:169.254.6.18. The postNAT Source IP will be169.254.7.10and the postNAT Dest IP will be169.254.6.18.
- Consider a flow which comes from LAN client with IP
Flow Link Stats Template
The Flow Link Stats template captures the flow stats broken down by link.
| Element ID (Enterprise Element ID) | Name | Type | Description | Applicable Edge Release |
|---|---|---|---|---|
| 148 | flowID | unsigned64 | An identifier of a flow that is unique within an observation domain. This information element can be used to distinguish between different flows if flow keys such as IP addresses and port numbers are not reported or are reported in separate records. | 3.3.0 |
| 1 | octetDeltaCount | unsigned64 | The number of octets since the previous report (if any) in incoming packets for this flow at the observation point. The number of octets includes IP header(s) and IP payload. | 3.3.0 |
| 2 | packetDeltaCount | unsigned64 | The number of incoming packets since the previous report (if any) for this flow at the observation point. | 3.3.0 |
| 32769 | octetDeltaCount_rev | unsigned64 | Biflow RFC5103. The number of outgoing bytes. | 3.3.0 |
| 32770 | packetDeltaCount_rev | unsigned64 | Biflow RFC5103. The number of outgoing packets. | 3.3.0 |
| 14 | egressInterface | unsigned32 | The index of the IP interface where packets of this flow are being sent. The value matches the value of managed object ifIndex as defined in [ RFC2863]. | 3.3.0 |
| 45008 (12240) | linkUUID | octetArray(16) | The internal link ID. | 3.3.0 |
| 45009 (12241) | replicatedPacketsRxDeltaCount | unsigned64 | Count of replicated packets received for the flow. | 3.3.2 (This field was part of template Id 256 in 3.3.0) |
| 45010 (12242) | replicatedPacketsTxDeltaCount | unsigned64 | Count of packets replicated for the flow. | 3.3.2 (This field was part of template Id 256 in 3.3.0) |
| 45012 (12244) | retransmittedPacketsTxDeltaCount | unsigned64 | Count of packets retransmitted for the flow. | 3.3.2 (This field was part of template Id 256 in 3.3.0) |
Tunnel Stats Template
A tunnel is established over a link and has communication with a peer. A peer can be a Gateway (edge to Cloud traffic), Hub (edge to data center traffic) or Edge (dynamic edge-to-edge VPN traffic). The Tunnel Stats template captures the stats of a tunnel and it is sent every one minute. The linkUUID field lists the link established for the tunnel. The interface Index field says to which peer it is communicating.
Difference between Tunnel and Path
Path is a unidirectional entity and tunnel is bi-directional. TX and RX paths make up a tunnel.
- Only connected tunnels will be exported. If a tunnel goes DEAD, this tunnel’s stats will not be exported from the next export interval. For example: if the tunnel stats template export interval is 300 seconds and the tunnel was exported at time t1 and tunnel goes down at t1+100. Stats between (t1 and t1+100) will be exported at t1+300. And from the next interval, this tunnel’s stats will not be exported since the tunnel has gone DEAD.
- Number of tunnels down events will be exported as part of tunnel stats template.
- Formula for Loss computation:
- TX Loss Percent = (( packetsLostDeltaTxCount) / ( packetsLostDeltaTxCount + packetsLostCompDeltaTxCount)) * 100
- RX Loss Percent = (( packetsLostDeltaRxCount) / ( packetsLostDeltaRxCount + packetsLostCompDeltaRxCount)) * 100
| Element ID | Name | Type | Description | Applicable Edge Release |
|---|---|---|---|---|
| 12 | destinationIPv4Address | Ipv4Address | This is destination Ipv4 address of tunnel. | 3.4.0 |
| 45008 (12240) | linkUUID | octetArray(16) | This is link UUID on which tunnel is established. This value points to entry in link option template (276). | 3.4.0 |
| 10 | interfaceIndex | Unsigned32 | This value identifies a peer. This value points to entry in interface option template (272). | 3.4.0 |
| 1 | octetsDeltaTxCount | Unsigned64 | Total bytes transmitted on this path. | 3.4.0 |
| 2 | packetsDeltaTxCount | Unsigned64 | Total packets transmitted out of this path. | 3.4.0 |
| 45079 (12311) | packetsLostDeltaTxCount | Unsigned64 | Total packets lost on this path. | 3.4.0 |
| 45083 (12315) | txLossPercent | Float32 | Loss percentage in this TX path. | 3.4.0 |
| 45058 (12290) | jitterTxMs | Unsigned32 | Tx average jitter of path in configured interval period. | 3.4.0 |
| 45060 (12292) | avgLatencyTxMs | Unsigned32 | Average TX latency of path in configured interval period. | 3.4.0 |
| 32769 | octetDeltaRxCount_rev | Unsigned64 | Total bytes received on this path. | 3.4.0 |
| 32770 | packetsDeltaRxCount_rev | Unsigned64 | Total packets received on this path. | 3.4.0 |
| 45011 (12243) | packetsLostDeltaRxCount | Unsigned64 | Total packets lost on this path. | 3.4.0 |
| 45084 (12316) | rxLossPercent | Float32 | Loss percentage in this RX path. | 3.4.0 |
| 45061 (12293) | jitterRxMs | Unsigned32 | RX average jitter of path in configured interval period. | 3.4.0 |
| 45063 (12295) | avgLatencyRxMs | Unsigned32 | Average RX latency of path in configured interval period. | 3.4.0 |
Application Option Template
The Application Option template is sent every five minutes or when changed. Only applications that have been referenced in flows are exported.
| Element ID | Name | Type | Description | Applicable Edge Release |
|---|---|---|---|---|
| 95 | applicationId | octetArray(8) | Scope field. Specifies an application ID. RFC 6759. | 3.3.0 |
| 96 | applicationName | string | Specifies the name of an application. | 3.3.0 |
| 372 | applicationCategoryName | string | An attribute that provides a first level categorization for each application ID. | 3.3.0 |
Application ID Format
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|20 | enterprise ID = 45346...|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|...Ent.ID.contd| app ID|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Classification Engine ID: 20 (PANA-L7-PEN)
- 45346 is VeloCloud SD-WAN PEN
- App ID is internal application ID
Interface Option Template
- Physical – These are Ethernet (e.g. GE1, GE2), VLAN (e.g. br-network1), or IP interfaces (e.g. PPPoE or some USB modem interfaces).
- SD-WAN – These are point-to-point interfaces between a pair of Arista devices. On the overlay, there may be several tunnels between a pair of Arista devices. These tunnels use a proprietary protocol called VCMP that provides several features including encryption, retransmission, and more. The tunnels between two devices may be always present or may be created on-demand depending on the configuration. The end points of these tunnels are called “links” in Arista terminology. Typically, there is a "link" for each physical WAN-facing interface on an Edge.
- Replicated on both the tunnels.
- Load-balanced between the two tunnels.
- Sent on only one tunnel.

The interface option template is sent every 5 minutes by default. The timer is configurable.
| Element ID (Enterprise Element ID) | Name | Type | Description | Applicable Edge Release |
|---|---|---|---|---|
| 10 | ingressInterface | unsigned32 | Scope field. The index of this interface. The value matches the value of managed object ifIndex as defined in [ RFC2863]. | 3.3.0 |
| 82 | interfaceName | string | A short name uniquely describing an interface, e.g. "Eth1/0". | 3.3.0 |
| 83 | interfaceDescription | string | The description of an interface, e.g. "FastEthernet1/0" or "ISP connection". | 3.3.0 |
| 45000 (12232) | interfaceType | unsigned8 |
|
3.3.0 |
| 45001 (12233) | destinationUUID | octetArray | Destination node UUID | 3.3.0 |
| 45013 (12245) | primaryIpv4Address | ipv4Address | Primary IP address of a physical interface. For SD-WAN interfaces this is always 0.0.0.0. | 3.3.0 |
Segment ID to Segment Mapping Template
The template is sent every 10 minutes and utilizes VRF as the nomenclature to define a segment.
Select
| Element ID | Name | Type | Description | Applicable Edge Release |
|---|---|---|---|---|
| 234 | ingressVRFID | unsigned32 | Scope field. A unique identifier of the VRFname where the packets of this flow are being received. This identifier is unique per metering process. | 3.3.0 |
| 236 | VRFname | string | The name of a VPN Routing and Forwarding table (VRF). | 3.3.0 |
Link Option Template
The link option template provides a mapping between linkUUID and the interface index to which this link points. From the link option template, it is also possible to get the link name which is a configurable field.
The Link Option template is sent every 5 minutes.
| Element ID (Enterprise Element ID) | Name | Type | Description | Applicable Edge Release |
|---|---|---|---|---|
| 45008 (12240) | linkUUID | octetArray(16) | The internal link ID. | 3.3.2 |
| 45078 (12310) | linkName | string | A short name uniquely describing the link. This is a configurable field in Orchestrator. | 3.3.2 |
| 10 | ingressInterface | unsigned32 | Index of underlying interface to which this link points. The value matches the value of managed object ifIndex as defined in [ RFC2863]. | 3.3.2 |
| 58 | vlanId | unsigned16 | The VLAN ID of this link. There can be more than one link on an interface which is differentiated by this VLAN ID. | 3.3.2 |
| 8 | sourceIP | unsigned32 | The source IP for this link. | 3.3.2 |
| 15 | nextHopIP | unsigned32 | The nextHop IP for this link. | 3.3.2 |
Netflow Source Address and Segmentation
Netflow source interface’s primary IP address should come from VeloCloud Orchestrator. In absence of the optional source interface configuration, the flow records would consume one of the up and advertised LAN/Routed IP address as source IP address. It is mandatory to have at least one up and advertised LAN/Routed interface on the particular segment, for Netflow to function. The Orchestrator UI needs to be modified to reflect this.
- Use different source interface for each segment.
- If we consider segments distinct exporting processes, then use observation DomainId to distinguish between segments.
Interface Mappings
0..7 0..7 0..16 destination_type reserved destination_if_idx
- E2E
- E2DC
- CLOUD
- ANY/DIRECT
- E2E, E2DC, CLOUD: map(next_hop_id) -> if_idx
- ANY/DIRECT: map(link_logical_id) -> if_idx
Filtering
- ingressVRFID (or all segments)
- ApplicationID
- sourceIPv4Address (mask)
- destinationIPv4Address (mask)
- protocolIdentifier
IPFIX Information Element Definitions
The following table lists the IPFIX information element definitions.
| 384 | valueDistributionMethod | A description of the method used to distribute the counters from contributing flows into the aggregated flow records described by an associated scope, generally a template. The method is deemed to apply to all the non-key information elements in the referenced scope for which value distribution is a valid operation. If the originalFlowsInitiated and/or originalFlowsCompleted information elements appear in the template, they are not subject to this distribution method, as they each infer their own distribution method. This is intended to be a complete set of possible value distribution methods; it is encoded as follows:
|
| 239 | biflowDirection | A description of the direction assignment method used to assign the Biflow Source and Destination. This Information Element may be present in a Flow Data Record or applied to all flows exported from an Exporting Process or Observation Domain using IPFIX Options. If this Information Element is not present in a Flow Record or associated with a Biflow via scope, it is assumed that the configuration of the direction assignment method is done out-of-band. when using IPFIX Options to apply this Information Element to all flows within an Observation Domain or from an Exporting Process, the Option must be sent reliably. If reliable transport is not available (that is, when using UDP), this Information Element must appear in each Flow Record. This field may take the following values:
|
Configure DNS Services
This is an optional service that allows you to create a configuration for DNS.
The DNS service can be a public DNS service or a private DNS service provided by your company. It is handled by the dnsmasq service, which sends the request to all the servers configured at the same time. The server with the fastest response is selected. The service is preconfigured to use Google and Open DNS servers.
Configure Private Network Names
You can define multiple private networks and assign them to individual private WAN overlays.
Configure Prefix Delegation Tags
Configure Authentication Services
If your organization uses a service for authentication or accounting, you can create a Network Service that specifies the IP address and ports for the service. This is a part of the 802.1x configuration process, which is configured in the profile.
Configure TACACS Services
TACACS services are used by organizations for authentication purpose to access the router or Network-attached Storage (NAS).
You can configure the TACACS settings for an Edge from the TACACS Services section available under category tab.
Configure Edge Services
This section allows you to configure VNFs and VNF Licenses. Virtual Network Functions (VNFs) are individual network services, such as routers and firewalls, running as software-only Virtual Machine (VM) instances on generic hardware.





































































