Mid- and large-size businesses are dumping aging PBX, Centrex, and keyphone telephony systems for Voice over IP (VoIP) faster than you can say “Ma Bell”. Cost savings can be significant, as can call quality, and the ability to manage the service from a web console is attractive to say the least. Monitoring and utilization tracking are usually quite simple and, most importantly, an increasing number of VoIP solutions are being integrated into unified communications (UC) offerings, allowing IT to roll out truly robust voice, video, presence, chat, and even email as a service.
But while the benefits are clear, many organizations stumble when they actually reach implementation because of infrastructure issues. VoIP-related infrastructure can be broken into four major areas:
- Networking hardware
- Bandwidth
- Directory services
- Monitoring
Although VoIP systems require their own hardware (e.g., SIP gateways and proxies, some of which may be virtualized or delivered as a service), the underlying networking hardware must be up to the task of providing high-bandwidth, low-latency connections required for clear, trouble-free calling. Adding other elements of unified communications only increases the requirements for fast hardware at both the edge of the network and throughout. There is nothing more frustrating than trying to carry on a sales call with any sort of latency affecting system performance.
Gateways, routers, and/or firewalls must not only be snappy with gigabit connections but must also be configured to prioritize voice traffic. Careful quality of service (QoS) implementations will ensure that critical voice connections aren’t affected by a lunch-time rush to watch YouTube. In addition to QoS and hardware considerations, a VoIP/UC rollout is a perfect time to review networking basics. Excessively long ethernet runs, cabling that can’t support gigabit speeds, or excessive hops to the network backbone can make for merely annoying delays loading a web page but turn into unacceptable quality when voice and video are flying over the wire.
Once internal infrastructure is sorted out, IT needs to look at bandwidth to the Internet. Is it fast enough to meet current and future needs? What options exist for increasing speeds, aggregating connections, and failing over to multiple providers to ensure service continuity? In the past, an Internet outage was a major problem but when that outage also takes down any voice communications, safety becomes as big an issue as productivity.
A VoIP rollout is also time to look at the state of directory services in an organization. Active Directory trees and other LDAP services have a tendency to sprawl and become overly complicated over time. User accounts don’t get cleaned up, policies become redundant or outdated, and groups get orphaned or ignored. Any VoIP system worth its weight will be integrated with directory services, so cleaning them up before rollout will make the process that much easier.
Finally, proactive monitoring systems need to be in place. Some of the monitoring functions can be automated through QoS, but it’s always best for IT to be able to identify network problems before users start to complain about dropped calls or poor call quality. Increases in jitter and latency or decreases in connection speed, among other metrics, can all be automatically reported to the right people in IT when they reach certain thresholds. What IT guy (or gal) doesn’t love telling a user reporting a problem that they’re already working on a fix?
Very few people will dispute the value of VoIP and the UC platforms of which it’s often a part. However, it can’t just be dropped into any functioning network and be expected to function perfectly. Infrastructure needs to be optimized before the first call ever gets made with the new system.
Chris Dawson is a writer, speaker, and analyst with particular interests in educational technology, healthcare IT, and the intersection of the two with the cloud and BI. He is a contributing editor at ZDNet, Ziff Davis, and UBM Channel, and a senior editor at Edukwest. You can follow him on Twitter (@mrdatahs) and Google+ (+Christopher Dawson).
