Prepping the network: 4 steps to collaboration

Deploying collaboration tools places new demands on the corporate network. Rollouts typically involve technologies such as Voice over Internet Protocol (VoIP) and others that require higher amounts of bandwidth and quality of service (QoS). Even if you already have VoIP in place, it's important to assess your network capabilities before adding on applications. An increase in traffic could result in the new tools or applications not working the way they should, which is likely to affect employee adoption rates, or at worst, vital communications systems could go down.

Here are four steps your organization should take to assess how prepared your network is for collaboration:

Step 1: Examine network and links documentation

Consolidating and examining documentation is the first stage of creating a picture of your organization's network environment. Documents should include:

Hardware specifications regarding switches, routers hubs and firewalls
Service provider reports, including security event monitoring information, bandwidth usage and network utilization by protocols
Software types, versions and upgrade/update paths followed

As you proceed with your assessment, refer back to the documentation frequently and consider whether it accurately reflects the actual state of the network. Any discrepancies should be properly documented, but they might also indicate ways your processes can be improved.

Step 2: Conduct site surveys

Before performing any network performance analyses, be sure to first address any physical or environmental problems. Walk around your sites and examine hardware and networking devices. A site survey should take into account power, HVAC, cabling, workstations, switches, routers and hubs, depending on LAN/WAN topology.

Consider, for example:

If cabling meets current needs, or if it's in need of replacement
If wiring closets are safely located, or is there potential for environmental damage due to a leaky water pipe or air conditioning unit, for example
If cabling topology follows accepted standards
If cabling is adequately labelled

In performing a LAN/WAN inspection, conduct a physical inspection as you did with cabling, followed by:

Logical (IP addressing) as well as physical inventories
A sample workstation analysis

A site survey should also take account of any known issues such as Wi-Fi "dead zones". It's important to conduct a physical assessment ahead of virtual analysis. An adequate physical infrastructure is a prerequisite for any systems testing. Any physical changes, like cabling, have the potential to delay your start date to a greater extent than any other issue.

Step 3: Test and analyze network performance

The only way that an organization can determine the potential for network congestion or other accompanying issues from a collaboration implementation is through monitored testing and analysis. This also establishes a baseline for measuring future performance.

A network performance assessment should typically analyze metrics in regards to:

Network devices–including switches, routers, hubs, etc. and capacity issues related to CPU, memory, buffer management, link/media utilization, and throughput
Applications–measuring for bandwidth and end-to-end delay, jitter and packet loss that email, VoIP, and file transfer protocol (FTP) might experience while traversing the network

In performing analyses, make use of tools such as packet/protocol analyzers, VoIP simulators and network diagram applications. These tools are indispensable for analyzing what happens to packets sent through different protocols. They also produce data that are used in statistical traffic analyses. Some tools generate network traffic for testing purposes, while others identify and record the characteristics of defective data packets, which helps determine sources of error.

The more powerful analyzers combine specialized hardware with high-performance analytical software. For your performance analysis, consider engaging a specialist who has invested in or developed hardware and software tools that can perform analyses detailed enough to pinpoint the causes of network error.

When conducting your performance evaluation, consider including these eight analyses:

Broadcast storm analysis–Isolates causes and effects of broadcast storms, which are a sequence of broadcast operations from a device or group of devices
Network capacity overload analysis–Examines network utilization and availability
Network throughput analysis–Measures actual speed of data transfer against standards
Network end-to-end inter-packet timing analysis–Measures timeliness of data transfer across the network
Transport and file retransmission analysis–Examines how often data packets are retransmitted; redundancy which may affect communication between workstations and servers
Packet route and path cost analysis–Examines the route of packet transfer throughout the network
End-to-end file transfer analysis–Examines file access process for any delays or interruptions caused by network delays or failure of an operating system, application or device
Trace Decoding–Examines data captured for any problematic packets or frames and isolates their causes. Data generated also provides an idea of what the network's optimum level of performance should look like

VoIP simulation testing is essential for collaboration implementations, as it determines the mean opinion score (MOS) of voice quality, taking total throughput, delay, jitter and packet loss into account. This helps to determine whether latency or bandwidth are issues.

Step 4: Perform a network threat risk assessment

Once you know what adjustments are necessary to accommodate new tools and increased traffic, review what you've discovered in the context of reducing risk. When the network is the vital link for nearly every kind of communication inside your organization, it is all the more critical that systems remain operational.

If you haven't already, consider performing a network threat risk assessment. Plan the assessment around three key concepts:

Redundancy: Are you ensuring multiple layers of redundant systems, such as having two routers and switches at each location?
Diversity: What options have you considered for load balancing?
Business continuity: How will you ensure that systems stay up in the face of any given scenario?

How you plan for potential failure scenarios depends on the architecture of your network and how you test your backup. Following best practices in building the network in the first place makes it easier to mitigate risk, and ultimately reduces subsequent costs of network modifications and risk mitigation.

Talk to Bell

Bell has significant experience in network collaboration assessment, planning, and upgrading.

A network assessment by Bell will help you determine whether remedial action is required in advance of any collaboration implementation. A Bell expert will:

Review your data and/or voice environment in terms of current functionality
Review network design based upon standards, current performance and ability to meet your requirements
Provide a baseline for future implementations