Ip Telephony Lifecycle

IP telephony lifecycle management White paper

Table of contents Introduction ................................................................................................ 2 Managing IP telephony: scoping the problem .................................................. 2 HP Management Solution for IP Telephony ..................................................... 5 Conclusion .................................................................................................. 8

This paper reviews the challenges of IP telephony, and then provides an overview of the HP IP Telephony Management Solution. This solution is based on a The adoption of IP telephony (IPT) is steadily growing lifecycle approach that optimizes the utilization of as companies recognize the value of converged voice people, process and technology for managing IPT. and data networks. Growing evidence today suggests From planning through delivery, to operation, the HP that IPT will eventually replace traditional telephony solution simplifies management of the complex IP teletechnology, and sooner rather than later. Helping drive phony environment and provides the highly specialized the success of IPT, early adopters are reaping the and automated management features needed for promised benefits of reduced operating costs, increased success. efficiency and productivity, and greater flexibility to deploy additional voice-related services.

Introduction

Of course, IP telephony is not without risk. Deploying IPT can involve significant commitments of capital and resources to replace a high-quality, reliable legacy phone system. In addition, once an IP telephony system is widely deployed, the phone service it provides is essential for conducting business and for the people who depend on it—people like customer service representatives, sales teams and stock traders, to name a few. Lastly, IP telephony systems are highly complex, realtime systems with stringent quality and availability requirements. Given the benefits and the risks of IP telephony, it’s vital that a company examines the challenges and employs a rigorous and holistic approach for deploying and managing IP telephony.

Managing IP telephony: scoping the problem The importance and complexity of an IP telephony system creates a number of challenges related to managing the environment. These challenges result from the fact that IP telephony: • Is a business-essential application and thus, drives stringent quality and availability requirements • Is one of the first real-time applications to be widely deployed on the IP network

• Is a service comprised of interdependent technologies including: applications, systems and special networking infrastructure • Is typically rolled out in phases over time

2

Figure 1. Network design considerations for VoIP

Endpoints

Access

Distribution

Core

WAN/PSTN

Network trust boundary

WAN WAN PSTN PSTN End-to-end delay should be less than 150ms

Sources of delay CODEC (G.729A 25 ms

Duplex mismatch (packet loss)

Queuing delay (variable)

Queuing delay (variable)

Jitter buffer (20–50 ms)

Serialization delay (variable) Propagation delay (6.3 ms/Km + variable network delay)

Network design considerations L2 QoS

Separate VLANs for voice and data

Rapid convergence

Rapid convergence

VLAN trunking Trust boundary

Rapid convergence

IP summarization

L3 QoS

L2 QoS

L3 QoS

Modularity WAN QoS Traffic shaping Link efficiency (LFI)

Source: Ramesh Kaza and Salman Asadullah, “Cisco IP Telephony: Planning, Design, Implementation, Operation, and Optimization,” Cisco Press, pages 117, 125, 2005

2.1 Stringent quality and availability

2.2 Real-time application

In general, a telephone service is considered a businessessential application. For example, voice communication is essential for discussion and correspondence among employees within a company, among a company and its vendors, and among a company and its customers. Thus, the availability and quality of the voice service can impact business performance, including revenue.

IP telephony is one of the first widely adopted real-time applications to run on the network and this presents some unique management challenges. As a real-time application, voice service quality is sensitive to network jitter, latency and packet loss. General thresholds for these metrics are as follows:

Next, in most cases, IP telephony is replacing a highly reliable and high-quality legacy phone service. And legacy phone systems have been known to work even during power outages. In order for an IP telephony deployment to be successful, it must meet or exceed these existing benchmarks. This translates to 99.999 percent (“five nines”) availability and “pin-drop” quality of service (QoS).

• Packet loss must be less than 1 percent

At the highest level, these availability and quality benchmarks drive the objectives for the management system. For example, an effective management system should provide comprehensive and detailed availability and quality reporting for the IP telephony system. Also, an effective management system should enable maximum availability and quality through reduced mean time to repair.

• Maximum jitter can be no higher than 40 milliseconds • Latency must be less than 100 milliseconds

To meet these stringent network performance criteria, it’s typical for an IP telephony deployment project to also include a significant amount of re-engineering of the IP network. Examples of network upgrades to support VoIP are provided in Figure 1. An effective management solution can play important roles in planning, delivering and supporting a realtime network for VoIP. For example, automatically configuring L2 and L3 QoS settings across a network can improve accuracy and save IT staff time. Furthermore, management solutions can be used to monitor and diagnose transient and hard-to-locate voice quality issues resulting from poor network performance and/or configurations issues.

3

Figure 2. Distributed IP telephony system

HQ

E-mail

IP PBX

Softphone

LDAP/AD IM etc.

FAX

PSTN

CO-Voice

CO-Data (WAN)

Internet VPN

LOC #1

LOC #2

2.3 Complex and multi-domain system IP telephony is a service comprised of interdependent technologies including: applications, systems and special networking infrastructure—all of which require management. Figure 2 and Table 1 provide a sample layout for a distributed IP telephony system and a listing of associated potential problems for the key components.

Table 1. IP telephony system component management requirements Component

Potential problems

IP PBX—call processing system services, application processes, system resources

Application availability Application process fault System overload

Gateway, gateway endpoints, channels

Overutilization Gateway endpoint shows availability but zero utilization

IP network

Availability and performance of IP network Routing configuration error sends voice traffic through inefficient path

IP phones

Poor quality during call

LOC #3

An effective management solution should manage across the entire IP telephony system from the phone, to the IP PBX, to the gateway and throughout the IP network. Furthermore, it should manage for service levels, faults, availability, performance, capacity, etc.

2.4 Phased rollouts Finally, within an enterprise, IP telephony is typically rolled out in a phased approach versus all at once and this drives special management considerations. There are several reasons for phased implementations of IPT including: mergers and acquisitions, depreciation schedules of legacy telephony systems, brand new sites and, probably most importantly, enterprises work to increase IPT deployment success by building a center of expertise around IP telephony deployments. Thus, a management solution should accommodate the lifecycle nature of IPT deployments. For example, the IPT management solution should be able to expand and scale as the IPT system grows overtime and it incorporates reusable templates (e.g. network configuration templates for QoS).

4

Figure 3. IP telephony management lifecycle

Plan

Deliver

Plan

Deliver

IPT business case/ROI

NW routing service design

Operate Fault, availability, diagnostics

Information and data gathering

QoS policy determination

Service level management

Network configuration

Performance management Capacity planning

Operate

HP Management Solution for IP Telephony The complexities of an IP telephony system and its phased deployment schedule create many challenges. To handle these technical and management challenges, it’s prudent to adopt a rigorous and holistic management approach. One such approach is provided by the IT Infrastructure Library (ITIL).1 ITIL provides an application lifecycle management model, which offers a best practice framework that accommodates people, processes and technology and is well suited for IP telephony. And because it’s a lifecycle model, it matches the phased rollout plans typical of IP telephony deployments. The HP IP Telephony Management Solution is based on the ITIL application lifecycle management model and is depicted in Figure 3. The following sections describe how the HP Management Solution for IP Telephony helps enable a successful IPT project at each stage of the lifecycle.

3.1 Plan The first step toward migrating to IP telephony is to work through the business case. Companies deploy IP telephony for economic and productivity benefits and an ROI analysis is a common first step. This paper assumes this step is complete. Once the business justification has been completed, the technology planning begins. The key objective of the technology-planning phase is to gather data for the delivery phase. So it’s important to look ahead to the delivery phase to understand what data is necessary to complete an effective design. Following are examples of types of information collected during the planning phase:

• Reliability of the existing network • Redundancy and resiliency of existing network • Capacity (bandwidth and ports) of existing network • Performance of existing wide area network (WAN) to handle IPT traffic The HP IP Telephony Management Solution provides several automated capabilities to aid in the planning phase. These capabilities include the following: • Network availability and fault reporting to determine the overall reliability of the network and individual components. This reporting enables the identification of network devices with the highest overall fault counts and the lowest mean time between failures, which are potential weak links within the network. • Automated network discovery and topology mapping. This can be used as the basis for network architecture analysis (i.e. subscription ratios between access, distribution and core layers). • Automated network inventory reporting. Inventory reports can be used to identify legacy network hardware and software to assess upgrades that will be required to support VoIP (e.g. identify which switches support L2 QoS and rapid convergence technologies). • Network performance reporting and capacity planning for establishing existing baseline capacity and performance. For example, performance reports can identify congested devices, interfaces and links. The Information Technology Infrastructure Library (ITIL) is a framework of best practice approaches intended to facilitate the delivery of high quality information technology (IT) services. ITIL outlines an extensive set of management procedures that are intended to support businesses in achieving both quality and value for money in IT operations. These procedures are supplierindependent and have been developed to provide guidance across the breadth of IT infrastructure, development and operations. [Source: Wikipedia]

1

5

The following example illustrates how HP software automates this process. First, the performance baseline is automatically captured from the existing network (see Figure 4 above). Then, the required VoIP bandwidth is estimated using a table similar to Table 2. Next, the estimated VoIP traffic load is injected into the network using an HP simulation and modeling engine. This modeling engine recalculates the routes and traffic across the network and develops reports showing the new state. See Figure 5. At this point, operators can change network infrastructure and configurations and re-run simulations to optimize the network to support the VoIP traffic.

Figure 4. Route analytic management system (RAMS) traffic view showing network performance baseline

3.2 Deliver The second phase of the IP telephony lifecycle is the delivery phase. The delivery phase includes both design and provisioning tasks. The design tasks utilize the information and data collected during the planning phase to design the IP telephony system and the underlying IP network to support VoIP. The provisioning tasks involve implementing and commissioning those designs. HP IP Telephony Management Solution provides capabilities to support the design and delivery tasks within the delivery phase. Design There are many aspects associated with designing an IP telephony system. However, this paper only focuses on some of the design tasks associated with the IP network to support VoIP that are aided by HP software. The first area of focus is associated with network configurations for VoIP. Because of the connectionless nature and required real-time performance of VoIP, it’s common to deploy special network configurations. Examples of configurations deployed to support VoIP include the following: separate VLANs for voice data, layer 2 and layer 3 quality of service and advanced router settings (e.g. link fragment interleaving). But in large multi-vendor environments, planning and design for network configurations can be a tedious, time consuming and error prone. Automated network change and configuration management software by HP provides a significant amount of automation and control for these tasks. Another key area of design for the network is the IP network routing service. Migrating from traditional telephony to IP telephony places additional traffic on the network and it must be designed for. HP provides automated simulation and modeling capabilities to aid in the design process for routed networks.

Provisioning Once the design phase is complete, the desired state must be provisioned throughout the network. Network changes and configurations must be carefully implemented, equipment must be commissioned and its proper functioning verified. This typically involves: • Deploying and configuring new hardware • Deploying new quality of service policies • Provisioning the IP telephony system (IP PBX, Gateways, etc.) HP IP Telephony Management Solution provides automation for several tasks associated with provisioning the IP telephony system. First, HP provides automated network change and configuration management for automating the configuration of network devices (e.g. L2 and L3 QoS configuration policies). Furthermore, HP WAN optimization capabilities enable real-time views of routing and traffic analysis for validating expected network performance. Table 2. VoIP bandwidth calculation example

Phones

Small site

Medium site

Large site

250

1,000

5,000

20%

20%

20%

50

200

1000

60%

60%

60%

Call profile Utilization Simultaneous calls Internal Internal calls

30

120

600

Intersite (LAN)

60%

60%

60%

Intrasite (WAN)

40%

40%

40%

Intersite calls (LAN)

18

72

360

Intrasite calls (WAN)

11

43

216

40%

40%

40%

20

80

400

3.2 Mbps

12.8 Mbps

64 Mbps

External PSTN External calls (PSTN) Utilization codec 6.711 (64 Kbps)

6

Figure 5. Network topology showing redesigned network links

Figure 6. Network node manager smart plug-in (NNM SPI) for IPT IP communication console

3.3 Operate The final phase in the IP telephony lifecycle is the operational phase. The objective of the operational phase is to maintain the desired state for the IP telephony system (i.e. to increase its availability and performance). Because an IP telephony system is a complex multidomain system a comprehensive end-to-end solution is required. That is, a solution that provides for the management of IP telephony services, applications, systems and networks. Furthermore, it’s critical that the HP IP Telephony Management Solution is integral with a company’s existing operational management framework to enable efficiency. HP provides a comprehensive solution for managing the operational phase of the IP telephony lifecycle, which includes the following: fault and availability management, performance management, capacity planning and service level management. Following are descriptions of each. Fault and availability management The HP solution provides for end-to-end IP telephony fault and availability management. The dashboard presented in Figure 6, illustrates the breadth of the HP solution. For example, it provides status indicators for IP telephony services (e.g. route patterns), IP PBX resources and performance monitoring statistics, gatekeepers and gateway endpoints, all in one view. From this one view, operators can quickly assess the health of their cross-domain IP telephony system and respond

efficiently. Also, each status indicator enables drilldown to underlying real-time metrics for additional information and links to performance reporting and capacity planning. In addition to overall monitoring of the IP telephony system, HP provides advanced diagnostic capabilities. Voice quality issues can be difficult to detect and diagnose. The HP solution is designed to catch voice quality issues in real-time (i.e. while they are occurring). The HP solution initiates when call metrics exceed performance thresholds and automatically analyzes the network path among phones to isolate the problem. An important benefit of this approach is that, once set up, it works automatically without operator involvement. Performance management Performance management is about verifying that the IP telephony system operates within acceptable performance thresholds on an ongoing basis and it’s about capturing data for diagnostics. The HP solution provides performance reporting and monitoring across an IP telephony system for the following: • CallManager system resources: CPU, memory, disk usage • CallManager usage/load: calls active, calls rejected due to call throttling • Call statistics: call success ratio (CSR), disconnect cause, call type distribution • Gateway, Gateway endpoints and Gateway endpoint channels: availability and usage

7

Figure 7. Automatic root-cause analysis for voice quality issues

NNM + IPT SPI

RCA Event SNMP Collect

SNMP Collect

x142 L2

L2

IPSLA Tests

WAN Problem between Nodes BLR_Edge and DEN_Edge, Phone: Src. Ext 132 (15.106.79.68) Dest. Ext. 142 (15.10.79.93)

HP Problem Diagnosis BLR_Edge Router

DEN_Edge Router

Figure 8. Voice quality service level reporting: mean opinion score (MOS) reporting

Capacity planning Capacity planning verifies that the IP telephony system is right-sized. Most of the data that is collected for fault and availability management and performance management is automatically trended and forecasted for capacity planning. For example, CallManager system resource metrics, such as CPU utilization, are trended and alerts can be generated when forecasted trends exceed acceptable capacity limits. Capacity planning is provided for the entire system including CallManagers, CallManager clusters, Gateways and Gateway endpoints. Service level management As with other business services, service level reporting is essential to provide ongoing service health and quality. Furthermore, service level reporting provides evidence and justification as to the success of an IP telephony project for stakeholders. The HP solution provides for automated service level reporting for service availability

To learn more, visit www.hp.com/go/software © Copyright 2007 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice. The only warranties for HP products and services are set forth in the express warranty statements accompanying such products and services. Nothing herein should be construed as constituting an additional warranty. HP shall not be liable for technical or editorial errors or omissions contained herein. 4AA1-6016ENW, October 2007

(voice services), CallManager availability and voice quality (i.e. minimum mean opinion score (MOS), average MOS and maximum MOS reporting).

Conclusion In summary, HP IP Telephony Management Solution provides the capabilities necessary to manage the IPT lifecycle and delivers: • Increased service availability and performance for the business-essential voice service • Reduced business risk and successful deployments of VoIP • Reduced capital and operational expense via optimized converged infrastructure • Increased operational efficiency and productivity for converged operations teams, engineering and telecom