The questions provided in this section will help you to gather detailed information on the customer's network infrastructure.
Network hierarchy and modularity are perhaps the two most important aspects of network design. A hierarchical network is easier to understand and easier to support because consistent, expected data flows for all applications occur on the network over similar access, distribution, and backbone layers. This reduces the overall management requirements of the network, increases understanding and supportability of the network, and often results in decreased traffic-flow problems, congestion issues, and troubleshooting requirements. Having a network hierarchy also improves the scalability of the network by allowing it to grow without requiring major network changes. Finally, a hierarchical network promotes address summarization, which is important in larger IP routing environments.
Table B-3 provides the questions that you should ask a customer to understand if the network design follows a hierarchical model.
No. | Question | Answer |
---|---|---|
1 | Does the network have separate core, distribution, and access layers that are appropriate for the number of campus users? | |
2 | Are users, servers, and WAN services connected to the access layer for all campus environments? | |
3 | If a distributed campus environment exists, does the network support a hierarchical and modular WAN core with defined core, distribution, and access connectivity? |
In this section, you need to collect the customer's LAN architecture diagram for all sites that require the IPT deployment. Based on the answers to the questions in Table B-3 and after studying the network layout, you need to document the strengths and weaknesses in the network and provide the recommendations to the customer to make the infrastructure ready to run IPT for this area of the network.
A modular network uses consistent network modules for the access, core, and distribution layers. Using a consistent "model" for each layer of the network improves the supportability because it becomes much easier to properly test modules, create troubleshooting procedures, document network components, train support staff, and quickly replace broken components.
Table B-4 provides the questions that you should ask a customer to understand if the network is designed following the modular approach guidelines.
No. | Question | Answer |
---|---|---|
1 | Does the network have consistent hardware and software modules deployed for the LAN server and user access? | |
2 | Does the network have consistent hardware and software modules deployed for the LAN distribution and core layers? | |
3 | Does the network have consistent hardware and software modules deployed for access and distribution layer WAN services? |
In this section, you need to obtain information about the router and switch hardware deployed at the core, distribution, and access layers. Use Table B-5 to gather the information, which will be helpful later to do the following:
Layer | Vendor, Product Family | Installed Modules | Part Number | IOS/CatOS Version Number |
---|---|---|---|---|
Core | Cisco Catalyst 6509 | 1000BASE-X Supervisor Multilayer Switch Feature Multilayer Switch Feature | WS-X6K-SUP1A-2GE WS-F6K-MSFC2 WS-X6348-RJ-45 | CatOS 8.1.(3) 12.1.20E2 |
Distribution | Cisco Catalyst 6509 | 1000BASE-X Supervisor Multilayer Switch Feature 10/100BASE-TX Ethernet | WS-X6K-SUP2-2GE WS-F6K-MSFC2 WS-X6348-RJ-45 | CatOS 8.1.(3) 12.1.20E2 |
Access | Cisco Catalyst 3550 | 48 inline power Ethernet 2 Gigabit Ethernet ports | WS-C3550-48-EMI | 12.1.20-EA1 |
Determine whether the routers and switches are capable of supporting inline power.
Determine whether the routers and switches have hardware or software revisions or versions that are required to run specific IPT features.
Calculate the number of inline power modules required to provide the power to IP phones.
Calculate the power supply requirements for modular switches.
This section looks at design and configuration considerations for the core layer for larger campus deployments. Smaller organizations might have more of a collapsed model, with the distribution and core layer functionalities combined in the same device.
Table B-6 provides a list of questions that you should ask a customer to understand the current core layer performance characteristics.
No. | Question | Answer |
---|---|---|
1 | Is it possible to increase the core capacity? | |
2 | Do core layer devices have the required system resources, including backplane utilization, packet forwarding capability, CPU and memory to support redundant core failover, and fast Layer 2 and Layer 3 convergence? | |
3 | If a Layer 2 core is used, will the Layer 3 neighbor count scale to the required growth? |
Describe the extent to which it is possible to increase the capacity of the core, either with existing hardware or by upgrading the hardware. Also, highlight any hardware limitations.
Table B-7 provides a list of questions that you should ask a customer to understand the current core layer high-availability characteristics.
No. | Question | Answer |
---|---|---|
1 | Does the campus core support redundant and modular core layer devices? | |
2 | Do Layer 3 equal-cost paths exist through the campus core for optimal routing convergence? | |
3 | Are core layer devices environmentally controlled and power protected for higher availability? |
Highlight any configuration problem that is not in line with best practices for the configuration of core switches.
Table B-8 provides a list of questions that you should ask a customer to understand the current core layer configuration characteristics.
No. | Question | Answer |
---|---|---|
1 | Are links between core layer devices and distribution layer devices hard configured for matching speed and duplex settings? | |
2 | If a Layer 3 core is used, are the interfaces configured with point-to-point subnets? | |
3 | Is ISL trunking or 802.1Q trunking eliminated from core layer devices? |
Highlight any configuration problem that is not in line with best practices for the configuration of core switches.
This section reviews the distribution layer architecture, high-availability considerations, and configuration of the distribution layer in a large campus environment. Smaller organizations might have a collapsed architecture. In smaller environments, for high availability, Cisco recommends combining the distribution and core layers and placing the servers on the access layer.
Table B-9 provides a list of questions that you should ask a customer to understand the current distribution layer performance characteristics.
No. | Question | Answer |
---|---|---|
1 | Does the distribution layer have increased bandwidth to handle access scalability, access aggregation, and distribution failover? | |
2 | Do the distribution layer devices have sufficient backplane, CPU, and memory resources for the required multilayer switching and features (including Hot Standby Routing Protocol [HSRP])? | |
3 | Do distribution layer devices have the required resources to support LAN campus QoS parameters deployed at the distribution layer? |
Highlight any configuration problem that is not in line with best practices for the configuration of distribution switches.
Table B-10 provides a list of questions that you should ask a customer to understand the current core layer high-availability characteristics.
No. | Question | Answer |
---|---|---|
1 | Does the distribution layer have redundant devices that can handle bandwidth requirements when alternate or primary distribution connectivity is unavailable? | |
2 | Has the organization anticipated failover and recovery scenarios with the chosen routing protocol, HSRP, and spanning-tree configuration? | |
3 | Is trunking between distribution layer devices configured only for VLANs where high-availability server access is needed for multiple access switches? |
Highlight any configuration problem that is not in line with best practices for configuration of distribution switches.
Use Table B-11 to collect the VLAN architecture information at distribution layer switches in the campus network.
No. | Question | Answer |
---|---|---|
1 | In general, are VLANs maintained on two distribution layer switches and one access layer switch only, and not carried on any link between the distribution layer switches? | |
2 | Are VLANs limited to two distribution layer devices and two access layer devices where high-availability access is needed? | |
3 | Are management VLANs separated from user VLANs at the distribution and access layers? |
Highlight any configuration problem that is not in line with best practices for configuration of distribution switches.
Table B-12 provides a list of questions that you should ask a customer to get an overview of the current distribution layer configuration.
No. | Question | Answer |
---|---|---|
1 | Is spanning tree configured on distribution switches, even when loops are not planned? | |
2 | Are the two distribution switches configured as alternating roots and secondary roots for alternating VLANs, with corresponding HSRP active and HSRP standby routers, for load balancing and redundancy? | |
3 | Has Backbone Fast been set for all distribution switches? | |
4 | Do links between access and distribution layer devices have hard-coded matching speed and full-duplex settings? | |
5 | Has trunk mode been set explicitly to on for trunking ports? | |
6 | Has channel mode been set to desirable for port channels? | |
7 | Is Uni-Directional Link Detection (UDLD) configured for switch-to-switch connections? |
Highlight any configuration problem that is not in line with best practices for configuration of distribution switches.
Table B-13 provides a list of questions that you should ask a customer to get an overview of current access layer characteristics.
No. | Question | Answer |
---|---|---|
1 | Does the access layer support 10/100-Mbps switched connections for all user end stations? | |
2 | Does the access layer support 100- or 1000-Mbps switched connections for all server and distribution connections? | |
3 | Does the access layer device support multiple queues to prioritize voice traffic where needed? | |
4 | Can the access layer devices handle peak utilization from servers and clients? |
Highlight your findings on the access layer characteristics.
Table B-14 provides a list of questions that you should ask a customer to understand the current access layer high-availability characteristics.
No. | Question | Answer |
---|---|---|
1 | Do access switches have redundant trunking to two distribution switches and only to the distribution switches? | |
2 | Do critical access switches have redundant processors and power supplies? | |
3 | Do high-availability servers have redundant connections to a single VLAN on two access switches? | |
4 | Is auto-negotiation configured for clients? | |
5 | Has duplex and speed been matched for server connections? | |
6 | Is PortFast configured on client/server ports? | |
7 | Is UplinkFast configured on switches that support high-availability servers, where forwarding and blocking VLAN trunks exist on the access switch? |
Highlight any configuration or design problems that are not in line with best practices for configuration of access layer switches.
This section looks at hub-and-spoke design considerations in Frame Relay hub-and-spoke environments.
Table B-15 provides a list of questions that you should ask a customer to understand how the WAN links are provisioned, current link utilizations and configured QoS parameters.
No. | Question | Answer |
---|---|---|
1 | Do you have a hub-and-spoke topology network? | |
2 | Will the WAN have redundant links to support high availability across the WAN? |
In this section, you need to obtain the following information pertaining to the customer's WAN architecture:
The customer's WAN diagram
A description of the overall WAN architecture
A list of the overall WAN strengths and weaknesses
This section examines whether the organization has properly planned for the addition of IP telephony traffic in terms of the existing traffic and the addition of RTP voice streams. This starts with a baseline of busy-hour or peak data traffic on trunk ports and extends to resource utilization for LAN devices, including buffers, memory, and CPU utilization on devices. Table B-16 provides a list of questions that you should ask the customer to gather the WAN base line information.
No. | Question | Answer |
---|---|---|
1 | Do you collect a baseline for your WAN (intended to support Cisco IP Telephony) that includes link utilization, queue depth, end-to-end packet delay, CPU, and memory? | |
2 | Has the organization determined the potential impact of Voice over IP (VoIP) traffic in the WAN in terms of bandwidth utilization and system resources? | |
3 | Does the current baseline and added traffic suggest that utilization and system resources are well within network capabilities, including peak link utilization below 75 percent and WAN links with a minimum of 128 kbps? |
Use this section to document the data for link utilization, end-to-end packet delay, and CPU usage for the WAN link. Also collect and document the information about memory that is installed on the router at the main location and at the branches.
Highlight any potential problems for IPT deployment.
You should ask the questions given in Table B-17 to analyze the capacity of the existing WAN and determine how much extra bandwidth is required to run the voice traffic along with the data traffic on the same links.
No. | Question | Answer |
---|---|---|
1 | Will the WAN be upgraded to support Cisco IPT solutions to help ensure consistent voice performance? | |
2 | Will WAN links that support voice traffic meet minimum bandwidth requirements for Cisco IPT deployments? (64-kbps frame, 64-kbps leased line, or 768-kbps ATM or ATM/Frame Relay is required.) | |
3 | In Frame Relay environments, will the organization have committed information rate (CIR) in the Frame Relay network for voice and configure traffic shaping to the CIR to guarantee voice traffic within the Frame Relay network? | |
4 | In ATM environments, will the organization use ATM traffic classes such as constant bit rate (CBR) or variable bit rate real time (VBR-rt) to guarantee critical voice traffic across the ATM network and shape ATM traffic to the guaranteed bandwidth? | |
5 | Will the WAN have adequate bandwidth to support peak voice usage across the WAN? |
Overall, is the network ready to run the voice traffic over the data network?
This section reviews the current IP addressing scheme and information on routing/routed protocols that are deployed in the data network.
Use Table B-18 to gather the information on the currently deployed IP addressing scheme for the data network and the customer's plans for assigning the IP addresses to IPT devices such as IP phones, voice gateways, call processing servers, etc.
No. | Question | Answer |
---|---|---|
1 | Provide the information on current IP addressing scheme. | |
2 | Does your organization have an IP addressing plan for integrating IP phones into the network? | |
3 | Does the organization have plans to implement RFC 1918 for private addressing? | |
4 | Does the IP addressing plan support IP address summarization? |
Use this section to document the currently deployed IP addressing scheme along with VLAN ID assignments, VLAN naming and numbering conventions followed.
Use table B-19 to analyze the current IP routing design and configurations deployed in the network.
No. | Question | Answer |
---|---|---|
1 | Has the organization implemented Open Shortest Path First (OSPF) or Enhanced Interior Gateway Routing Protocol (EIGRP) for improved network convergence? | |
2 | Does the organization use any other routing protocols that redistribute to/from the primary EIGRP or OSPF autonomous system? | |
3 | Does the organization maintain standard routing protocol configurations for all routers in the network? | |
4 | Are static routes confined to network edges for partner connectivity or ISDN backup? | |
5 | Has the organization implemented IP summarization toward the core to reduce routing protocol overhead and ensure IP scalability? | |
6 | Has the organization implemented stub or default routing in WAN hub-and-spoke environments to reduce routing protocol traffic overhead on WAN links? | |
7 | Has the organization reviewed routing protocol impact and scalability based on device types, number of routes, and IP routing protocol neighbors? | |
8 | Is routing disabled on user and server LAN interfaces to prevent core routing through user LANs? | |
9 | Is routing filtered on access site WAN interfaces to advertise only WAN site information? | |
10 | What other interior routing protocols are used in the network other than IP? | List: |
Describe the overall routing design.
Provide network maps and diagrams for the Layer 3 design.
This section looks at HSRP considerations with respect to the common infrastructure model. Table B-20 provides the questions that you should ask a customer to understand how the HSRP is designed in the network.
No. | Question | Answer |
---|---|---|
1 | Does the network use HSRP for redundant default gateway support? | |
2 | Does the organization understand convergence issues given the number of HSRP groups supported on the device? | |
3 | Does the network use the HSRP preempt feature to return control to the primary gateway, closely associated with the spanning-tree root for the VLAN? | |
4 | Has the organization considered the HSRP track feature that is used to track backbone or WAN connectivity from the primary HSRP gateway? |
Describe how and where HSRP is deployed in each part of the network. HSRP design might be complex when it is deployed in Layer 2 switches with internal and chassis redundancy.
You should use the questions provided in Table B-21 to analyze whether the network devices have the quality of service (QoS) capabilities to support IPT.
No. | Question | Answer |
---|---|---|
1 | Can auxiliary VLANs with 802.1Q/p be used for voice? | |
2 | Can the voice bearer streams be marked as EF (Expedited Forwarding) and voice control streams as AF31 (Assured Forwarding 31)/CS3 (Class Selector 3)? | |
3 | Is QoS configurable on devices where buffers might be reaching capacity? | |
4 | Is Low Latency Queuing (LLQ) configurable on all WAN interfaces? | |
5 | Is Link Fragmentation and Interleaving (LFI) configurable on all links where speed is below 768 kbps? | |
6 | Do you have any mission-critical traffic besides voice (e.g., video, DLSW, etc.)? If Yes, please specify the traffic types and how they are currently classified in your network. |
Describe the QoS strategy in place and highlight any potential problems when deploying voice over the data network.