Nodes and Sites
Both node and site are WAE Design terminology.
- Node —A device in the network, which can be one of three types: physical, PSN (pseudonode), or virtual. The Type property distinguishes whether the node represents a real device or an abstraction, such as a single node representing a number of edge nodes connected in the same way to the network. A physical node is a Layer 3 device, or router. Physical and virtual nodes behave in the same way within WAE Design. A pseudo-node (PSN) is typically used to represent a Layer 2 device or a LAN.
Nodes can reside both inside and outside of sites. The external arrangement could be useful for small networks where routers are not geographically dispersed.
- Site —A collection of nodes and/or other sites that potentially form a hierarchy of sites. Any site that contains other sites is called the parent site.
Both nodes and sites have simulated traffic, while nodes also have measured traffic.
- Nodes are shown as rectangles. The border color is an indication of traffic sourced from and destined to the node.
- Sites are shown as squares. The border color indicates the traffic utilization of all the nodes and circuits inside the site, including all nested nodes and circuits.
Note that sites can contain both L3 nodes and L1 nodes. Empty sites appear in both views, as do sites containing both L3 and L1 nodes. Only sites containing L3 nodes appear in the L3 view, and only sites containing L1 nodes appear in the L1 view. To change these defaults, see the WAE Network Visualization Guide.
Parent Sites and Contained Objects
Unless a site is empty, there is a hierarchy of the sites and nodes that a site contains. A site can be both a parent and a child site. If a site contains another site, it is a parent site. Any nodes, sites, or circuits within a parent site are contained ( nested) objects. The contained nodes and sites are also called children. Often the child nodes and sites are geographically co-located. For instance, a site might be a PoP where the routers reside.
- The site’s Parent Site property defines whether it is nested within another site. If it is empty, the site is not nested.
- In the network plot, the parent site shows all egress inter-site interfaces of all nodes contained within it, no matter how deeply the nodes are nested. Similarly, this is true for each child site plot within it.
- Selecting a site from the network plot does not select the sites or nodes under it.
- To filter to the sites or nodes contained within a site, right-click to access its “Filter to contained” menu. If you select “Direct,” WAE Design filters to only those objects contained immediately within it. If you select “All,” WAE Design filters to all objects contained within it, no matter how deeply they are nested.
Figure 3-1 shows an example of nodes outside of sites (atl nodes), sites containing other sites (the East site contains three NYC sites and one BOS site), and nodes within a site (NYC_2 site contains three nyc nodes). While this figure only shows one of the child site plots, if you were to expand all nested sites with the East site, you would find more nested child sites and nodes. As you can see in the network plot, the East site has 10 egress interfaces extending from it, some of which are within nested sites.
Figure 3-1 Example of Node and Site Relationships
Selecting a site from the network plot does not select the sites or nodes under it except when deleting the site. In this case, all objects within a site are selected for deletion. However, in the confirmation that appears, you have the option to keep the contained sites and nodes. If you do, then the objects that are contained directly within it are moved to be on the same level as the site that is removed. The other, more deeply nested objects maintain their parent relationships.
Figure 3-2 shows a West site plot directly containing two sites (SJC and LAX) and three nodes. This parent West site resides in the network plot. Both SJC and LAX contain two sites each. By deleting the West site plot, but retaining its children, both SJC and LAX move to the network plot level, as do the three sea nodes. The sites within SJC and LAX remain as children sites to their respective parent sites.
Figure 3-2 Example Site Deletion
WAE Design plan files can contain nodes of with a Type property of "psn." These nodes represent pseudonodes (PSNs), which are used to model LANs or switches that connect together more than two routers. They are used in two situations: for IGP modeling and BGP peer modeling.
In an IGP network, a LAN interconnecting multiple routers is represented by a PSN node, with circuits connected to each of the nodes representing the interconnected routers. Both OSPF and IS-IS have a built-in system whereby one of the routers on that LAN is the designated router (DR) for OSPF or the designated intermediate system (DIS) for IS-IS. The PSN node is named after this designated router. WAE Collector creates nodes with a property Type of psn automatically during IGP discovery.
When BGP peers are discovered, WAE Collector might find that a router is connected to multiple peers using a single interface. This is typical at switched Internet Exchange Points (IXPs). WAE Collector then creates a node with a property Type of psn, and connects all the peers to it, each on a different interface.
Following are a few points to consider when working with nodes that have psn as a Type property.
- Two PSNs cannot be connected by a circuit.
- If a PSN node is created by WAE Collector, “psn” is prepended to the designated router’s node name.
- When creating demand meshes, WAE Design does not create demands with nodes of Type psn as sources or destinations. This is possible in manual demand creation, but not recommended. WAE Design sets the IGP metric for all egress interfaces from a node of Type psn to zero. This ensures that the presence of a PSN in a route does not add to the IGP length of the path.
To create a node, either select the Insert->Node menu or right-click in an empty plot area and select New->Node from the context menu. Following are some of the frequently used fields.
- Name—Required unique name for the node.
- IP Address—Often the loopback address used for the router ID.
- Site—Name of the site in which the node exists. If left empty, the node resides in the network plot. This offers a convenient way to create a site while creating the node, move nodes from one site to another, or remove nodes from a site so it stands alone in the network plot.
- AS—Name of the AS in which this node resides, which identifies its routing policy. This can be left empty if no BGP is being simulated.
- BGP ID—IP address that is used for BGP.
- Function—Identifies whether this is a core or edge node.
- Type—The node type, which is physical, PSN, or virtual. Because a PSN node represents a Layer 2 device or a LAN, interfaces on a PSN must all have their IGP metrics set to zero, and two PSN nodes cannot be directly connected to one another. If you change a node type to PSN, WAE Design automatically changes the IGP metrics on its associated interfaces to zero.
- Longitude and Latitude—Geographic location of the node within the network plot. These values are relevant when using geographic backgrounds.
- X and Y—Location of the node within the network plot. These values are relevant when using schematic backgrounds.
To create a site, either select the Insert->Site menu or right-click in an empty plot area and select New->Site from the context menu. Following are some of the frequently used fields.
- Name—Required unique name for the site.
- Display Name—Site name that appears in the plot. If this field is empty, the Name entry is used.
- Parent Site—The site that immediately contains this site. If empty, the site is not contained within another one.
- Location—To automatically place a site in its correct geographic location and update the Longitude and Latitude fields, enter the airport code and press Enter. To select from a list of cities worldwide, press Alt-Enter.
- Longitude and Latitude—Geographic location of the site within the network plot. These values are relevant when using geographic backgrounds.
- Screen X and Y—Location of the site within the network plot. These values are relevant when using schematic backgrounds.
Duplicating a node creates an identical node and duplicates the connectivity of that node by creating new, identical circuits containing the same amount of traffic. By default, each demand and each LSP that uses the node as a source or destination is also duplicated. The newly created node is sequentially named the same. For example, if you duplicated a node named “chi,” WAE Design would create a chi node. If you were to duplicate it again, a chi node would be created.
Step 1 Use one of these methods to access the Duplicate Node dialog box.
- Right-click a node, and select Duplicate from the context menu.
- Select a node, and select the Edit->Duplicate menu.
Step 2 Select whether to duplicate demands and LSPs, and then click OK.
Duplicating a site creates an identical site, including the nodes within it and its connectivity. The default site name and display name are the same as the copied site plus a sequential number. For example, duplicating OKC would create OKC. However, you have the option to change both the site name and the display name. The nodes that are created are likewise sequentially renamed. To change a duplicated node name, open the node’s Properties dialog box.
Note You cannot duplicate a site that contains other sites.
Step 1 Use one of these methods to access the Duplicate Site dialog box.
- Right-click a site, and select Duplicate from the context menu.
- Select a site, and select the Edit->Duplicate menu.
Step 2 If desired, rename the new site and/or display name, and click OK.
Real network topologies often have a number of nodes, typically edge nodes, connected to the network in the same way. For example, they might all be connected to the same core node or pair of core nodes. For planning and design of the network core, it is often desirable to merge these physical nodes into a single virtual node, which simplifies the plan and accelerates the calculations and simulations performed. Note that merging nodes changes the plan itself, not just the visual representation.
The name of a newly merged node can be based on the site name, selected node name (base node), or have a new user-specified name. The node merge effects are as follows.
- Re-attach circuits from other nodes to the base node.
- Move demands to or from other nodes to the base node.
- Move LSPs to or from other nodes to the base node.
- Set base node traffic measurements to the sum of measurements of the selected nodes.
- Delete other nodes.
Step 1 Select multiple nodes. If you do not select nodes, WAE Design merges all nodes.
Step 2 Either select the Initializers->Merge Nodes menu, or right-click one of the nodes and select Merge from the context menu.
Step 3 Select whether to merge the nodes per site or merge them into one node.
- Separate merge per site—Merges nodes on a per-site basis. For example, if you selected all nodes in the plan, the result would be one merged node per site. If you do not specify a new suffix, the default name is the same as the site.
- Merge all nodes together into base node—Merges all nodes selected into one node. For example, if you selected two nodes in one site and three nodes in another, the result would be a single node in the site and node combination selected as the base node.
If you do not specify a new name, the default is to use the name of the base node.
Step 4 Click OK.
Trim Nodes and Move Their Traffic
A plan file created through WAE Collector discovery often has more edge nodes in the plot than you are interested in viewing or simulating. These might be nodes that forward very little traffic or are singly connected to the core. However, you can reduce such a network model in WAE Design for simplicity and performance improvements by trimming (removing) these nodes. This process also removes all associated circuits, demands, and external endpoints.
Note that you can either select nodes to trim or you can choose to trim all nodes except those that are selected.
Typically, plan files are trimmed before demands are created. However, you can optionally trim demands in a plan file. This moves a demand that is sourced on a trimmed node to the first hop on the demand path in the remaining network. Likewise, a destination to a trimmed node is moved to the last hop of the demand path in the remaining network. If two or more demands in the same service class are trimmed so that their resulting source and destination nodes are equal, these demands are aggregated into one demand with their traffic summed. Demands split by ECMPs are converted into multiple demands, each with traffic divided proportionally to the ECMP split.
Trimming a node is not the same as deleting a node. When you trim a node, the measured traffic carried through that node is transferred to the nodes that remain so that the overall traffic in the network remains unchanged. If the node is deleted, this traffic is discarded. Only measured traffic is preserved in this way.
Note that multicast demands are removed, and cannot be trimmed.
Example: Edge node E is singly connected to node C in the core. When E is trimmed, the measured traffic on the interface from E to C is added to the total source traffic on node C. The measured traffic on the interface from C to E is added to the total destination traffic on node C. By default, demands to and from node E are removed.
Example: A demand source is node A, and destination is node E, and is routed through nodes A-B-C-D-E. If nodes A and E are trimmed and if you select the option to trim demands, then the new route for the demand is B-C-D.
Step 1 Optional: If you want to trim the traffic of nodes under a specific utilization, sort one of the Traffic columns to find the lowest to highest utilization. Alternatively, use the Nodes table filter to search for traffic of a certain level using, for example, the less than operator (<).
Step 2 Select one or more nodes.
Step 3 Right-click and select Trim from the context menu.
Step 4 Optional: Select the option to trim demands.
By default, the moved demands are named “Trimmed,” though you can enter a different name.
Step 5 Click Yes to confirm the action.
Circuits and Interfaces
In WAE Design, an interface is a either an individual logical interface or a LAG logical interface. If there is a one-to-one mapping between a logical and physical interface, then the interface contains both Layer 3 properties (for example, Metric) and physical properties (for example, Capacity). If there is a one-to-many mapping between logical and physical interfaces, then the interface is the logical LAG and the ports are included in the plan file as the physical ports in the LAG. For more information on ports and port circuits, see the Ports, Port Circuits, and LAGs section.
Each circuit connects a pair of interfaces on two different nodes. Therefore, an interface always has an associated circuit. Both the interface Properties dialog box and the circuit Properties dialog box enable you to simultaneously edit properties for the pair of interfaces and the circuit.
Interfaces have both measured and simulated traffic. The traffic that appears in the Circuits table is the higher of the traffic in the two interfaces.
The Latency and Distance Initializer uses mapped L1 circuit delays to calculate L3 circuit delays. If L1 circuits are not mapped to the L3 circuits, node Longitude and Latitude properties are used in the calculation. For information, see the Simulation chapter.
Create Circuits and Interfaces
To create a circuit, either select the Insert->Circuit menu or right-click in an empty plot area and select New->Circuit from the context menu. Following are some of the frequently used fields, with the critical ones being the definition of the interfaces and the capacity.
As a result of creating a circuit, two interfaces are also created.
- Capacity—The amount of total traffic this circuit can carry. The drop-down list has a selection of the most widely used capacities.
- SRLGs—If you want this circuit to belong to an SRLG, select it from this list or create a new one by clicking the Edit option. See the Create SRLGs for Circuits Only section for more information.
- Parallel Group Name—To include this circuit in a new or existing parallel grouping, enter its name. To visually combine these parallel groups in the network plot, see the WAE Network Visualization Guide.
- Interface A and B—You must specify two interfaces that are connected by the circuit.
Duplicating a circuit has the following effects on the new circuit and its interfaces.
- Duplicates the circuit itself and all its properties, including the circuit name.
- As a result, it duplicates both associated interfaces and renames them by adding an incremental numeric suffix to the original names, such as . All other interface properties are copied to the new interfaces.
Step 1 Select one or more circuits to duplicate.
Step 2 Right-click one of these circuits, and select Duplicate.
Step 3 Enter the number of new circuits you want to create for each selected circuit. For example, if you selected 2 circuits and you entered a 3, each of those selected circuits would have 3 new parallel circuits for a total of 6 new circuits.
Step 4 To duplicate L1 circuits that are mapped to selected L3 circuits, select “Duplicate mapped L1 Circuits.”
Step 5 Click OK.
Create a Circuit Mesh Between Nodes
To create a mesh of circuits between multiple nodes, follow these steps. Figure 3-3 shows an example circuit mesh created between all the nodes within a site.
Step 1 Select multiple nodes that you want to interconnect.
Step 2 Right-click the interface you want to duplicate, and select Duplicate Between Selected Nodes from the context menu.
Step 3 Click OK to confirm.
Figure 3-3 Example Circuit Mesh
You can simplify the plan by merging circuits that have the same source and destination endpoints (nodes). This capability is useful, for example, in long-term capacity planning where multiple parallel circuits can be ignored and only the site-to-site connections are of interest.
Merging circuits changes the plan itself, not just the visual representation. (To change only the visual representation, see the WAE Network Visualization Guide.).
The circuit merge effects are as follows.
- Set base circuit capacity to the sum of capacities.
- Set base circuit metric to minimum of metrics.
- Set base traffic measurements to the sum of measurements.
- Delete other circuits.
Step 1 Select multiple circuits that have the same source and destination endpoints.
Step 2 Either select the Initializers->Merge Circuits menu, or right-click one of the circuits and select Merge from the context menu.
Step 3 The name of a newly merged circuit is based on the selected circuit. If you would like to base the name on a different circuit, select it from the Base Circuit list.
Step 4 Click OK.
Ports, Port Circuits, and LAGs
In WAE Design, a port is a physical interface. You can model LAGsFigure 3-4).
By default, each logical interface listed in the Interfaces table corresponds to a single physical port, and these ports need not be explicitly modeled. The exception is when the logical interface is a LAG, which bundles more than one physical port. In this case, the physical ports are listed in the Ports table.
A port circuit is a connection between two ports. However, ports are not required to be connected to other ports by port circuits.
Port circuits can be mapped to L1 circuits, and ports can be connected to L1 nodes. For information, see the Layer 1 Simulation chapter.
Figure 3-4 Ports, Port Circuits, and LAGs
To create a port, either select the Insert->Port menu or right-click in an empty plot area and select New->Port from the context menu. Following are some of the frequently used fields.
- Site and Node—Site and node on which this port exists.
- Interface—Logical interface to which this port is mapped. This must be defined to create port circuits using this port.
- Capacity—The amount of total traffic this port can carry. The drop-down list has a selection of the most widely used capacities.
Each port can be connected to an L1 node. Creating these connections populates the L1 Node column in the Ports table. If this port belongs to a port circuit, the Remote L1 Node column is also updated. To connect the port to an L1 node, select the Layer 1 tab. Then select the site containing the L1 node and also the L1 node to which this port is connected. For more information on Layer 1, see the Layer 1 Simulation chapter.
To make one of the existing interfaces into a LAG by assigning ports to it, follow these steps. If the interface does not contain ports, you must first create them (see Create Ports).
Step 1 From the Ports table, select all ports that belong to the LAG. You can quickly do this by filtering to their common interface.
Step 2 Double-click one of the ports to open the Properties dialog box.
a. Select the interface that contains these ports.
b. Click OK.
Create Port Circuits
A port circuit specifies a pair of connected ports.
- Both of the ports must exist and be mapped to interfaces that are connected by a circuit.
- When selecting the two ports for the port circuit, note that if one is assigned to an interface, the other must be assigned to the remote interface on the same circuit.
To create ports and map them to their interfaces, see Create Ports.
To create a port circuit, either select the Insert->Port Circuit menu or right-click in an empty plot area and select New->Port Circuit from the context menu. The site, node, and port are all required fields.
- Site and Node—Site and node on which the port exists.
- Port—Name of the port.
- Capacity—The amount of total traffic this port circuit can carry. The drop-down list has a selection of the most widely used capacities.
Each port circuit can be mapped to an L1 circuit. Creating these mappings populates the L1 Circuit Name, L1 Circuit Node A, and L1 Circuit Node B columns in the Port Circuits table. For more information on Layer 1, see the Layer 1 Simulation chapter.
Step 1 Select the Layer 1 tab.
Step 2 Select the sites and the L1 nodes within them on both ends of the L1 circuit.
Step 3 Either map an existing L1 circuit or create a new one, and then click OK.
- To map an existing L1 circuit, click Select and then select the L1 circuit from the drop-down list.
- To create an L1 circuit, click Create. Either enter a new L1 circuit name or use the default one that is shown. Once you click OK, you have the option to create L3-L1 links.
Set LAG Simulation Properties
Each interface is considered to be a LAG (port channel). You can configure LAG properties so that if it loses too much capacity due to inoperational ports, the entire LAG will be taken down.
Step 1 Double-click an interface or circuit; the Properties dialog box appears.
Step 2 Select the Advanced tab and go to the Port Channel section. You can set one or both of the following parameters, and you can set them for either one or both interfaces on the circuit.
a. In the Min Number of Ports field, enter the minimum number of ports that must be active and operating for the LAG circuit to be up.
b. In the Min Capacity field, enter the minimum capacity that must be available for the LAG circuit to be up.
Step 3 Click OK.
Many objects can be tagged as a way to group them for subsequent processing. Once created, tags appear as options in dialog boxes, enabling you to select and act upon all objects with a given tag name. Thus, tags provide a very flexible way to group interfaces, demands, nodes, LSPs, sites, and AS’s that are not easily selected using standard attributes. For example, when using LSP paths you might want to tag selected paths to make it easier to assign or change their path affinities.
Adding tags to a parent site does not add tags to its children sites and nodes.
Step 1 Right-click one or more objects and select its Properties menu.
Step 2 Select the Tags tab.
Step 3 Click New Tag; a New Tag dialog box appears.
a. Enter a name that identifies the objects you are tagging.
b. Click OK.
Step 4 Click OK.
Enable and Disable Tags
Step 1 Right-click one or more objects and select its Properties menu.
Step 2 Select the Tags tab; a list of available tags appears.
Step 3 Toggle the tags to enable (T = true) or disable (F = false).
Step 4 Click OK.