Is an IP Telephony system different from other network applications?
While the IP telephony system may operate over an existing IT network, it requires a higher degree of reliability. Most end users see the telephone as a “bulletproof” system, expecting it to work even when nothing else does. To match the reliability of standard phone systems, an IP telephony system must have extended battery back-up or generator support, true-online, double-conversion UPS on every router, server, switch and other important component, and the capability to power individual phone sets through the communications wires (a capability that standard phone systems possess today). The need for connecting to 911 services even when there is a power outage makes online protection even more critical.
Is power conditioning as important as battery back-up?
In a word, Yes. This is even more important with the newer, gigabit-scale routers and switches. While the average network must deal with about 15 significant power outages a year – 90 percent of which last less than five minutes – most networks face an average of one power disturbance per day. Some network managers have already found out the hard way that anything but true online, double-conversion UPSs can let power problems through to switches and routers. Line-interactive UPS equipment loses contact with ports, drops packets, and even shuts down unexpectedly when power disturbances pass through the UPS. It can cause problems merely by switching to and from battery to “protect” vital systems. If applications are mission critical, the only safe route to high availability is an online, double-conversion UPS.
Is the IP Telephony system being added to an existing network?
If so, make sure your protection is up to the “Five Nines” standard defined by Cisco and other equipment providers. Many rack-based networks are operating at a lesser standard because they have line-interactive protection, and many others may lack the back-up power capacity to provide a phone system protection that meets user expectations. Adding more load or longer ride-through times may require a shift from rack-mounted protection to room-level protection. Now is the time to re-think your protection scheme, and build high availability into all mission-critical applications.
What does the industry say about maintaining high availability for IP Telephony systems?
The Cisco white paper, IP Telephony: The Five Nines Story, states that high-availability protection – defined as “five-nines” (about 5 minutes of downtime per year) – must include “UPS and generator backup for all distribution, core, gateway and Cisco CallManager devices.”
The paper goes on to list Cisco’s high-availability requirement:
- UPS and generator backup
- UPS systems have auto-restart capability
- UPS system monitoring
- Four-hour response service contract for UPS system problems
- Maintain recommended equipment operating temperature 24x7
This has been Liebert’s basic standard for high availability for decades, and reflects Liebert’s long-standing product and service mix.
Where is the IP Telephony equipment being installed?
Even when other network applications are centralized in a protected datacenter, an IP telephony system is likely to have at least some decentralized pieces of vital hardware. Adding a router or switch to a server stack implies one kind of protection need. Installing an IP telephony application in a campus environment requires high-availability protection for every node in the network, including, in many cases, desktop phone sets and wireless LANs. Power protection becomes something far more complex and challenging than the approach used for traditional IT applications.