|
 By Dr. Jim Kennedy, NCE, MRP, MBCI, CBRM
Over the last few years small, medium, and large businesses have begun the revolutionary movement from conventional telephone services to VoIP (aka Voice over IP). They have either begun to utilize the many VoIP service offerings made by new or existing carriers or have designed and developed a proprietary VoIP infrastructure within their organizations.
With this movement towards VoIP networks accelerating at an ever increasing rate there is a greater concern about maintaining its high level of availability to that which we have come to rely upon for traditional telephony services.
In order to accomplish this, the concern has focused on business continuity and availability of VoIP services. Ultimately the business continuity and/or availability models or methodologies that will be chosen for VoIP will depend on business models and technical implementation of infrastructure and services.
Understanding VoIP
For the most part everyone is familiar with the intricacies and capabilities of the traditional telephone services. Changing over to VoIP is changing to a technology that is constantly being benchmarked, even if in a single individual’s mind, against that level of service and quality. It is also important to realize that even though it is benchmarked against public utility telephony services it is still a convergence of voice over a data network. Some of the VoIP requirements are similar to data IP networks, but there are several requirements that VoIP relies upon to meet users’ expectations that exceed typical data networking standards.
VoIP requires real-time service
In order for VoIP to maintain its high level of quality all conversations need to be accomplished without any network delays. This means that any automatic fail-over or dynamic reconvergence of networks around failed components needs to be done in real-time. Otherwise a conversation could be compromised.
Hardware redundancy
In many cases organizations implement VoIP within their operations, not relying upon outside services or carriers. In this case there are many different components and applications that can be selected to provide for the organization’s telephony needs. When designed correctly these components are integrated in a redundant fashion to ensure fail-over and high availability. If they are not, then the high level of availability which traditional telephony services have gained a reputation for will not exist and failures of telephone operations will most certainly occur.
Network capacity
Since the quality of the telephone conversation is reliant on the fact that packets of information are transmitted and received on a real-time basis it is important that any designs include the capability of getting those packets through even during times of high usage by other voice conversations and data communications needs. If this is not the case then conversations could appear choppy or even distorted with portions of a conversation lost in the network abyss.
Business continuity and VoIP
So you are probably beginning to ask yourself by now what does all this have to do with business continuity and VoIP?
Good question. We have become to rely upon the perceived fact that the phones will always be there when we need to use them even during times of emergency and crisis. With VoIP this can be true if VoIP is properly engineered, implemented, and planned. If you use a service provider to supply VoIP services the phone will be there if you have done your due diligence to make sure that your provider meets your expectations for phone service availability. With VoIP the high availability capability does not automatically come with the technology or service - it needs to be designed in or planned for.
There are several things that need to be considered when building and/or choosing a VoIP capability within an organization:
Power
One area where VoIP differs from traditional phone service is that the VoIP phones on the desktop are powered by your organization not the phone company. So, if in your business continuity plan you have provided for UPS (uninterruptible power supply) support for your network you need to take a special look to ensure, if VoIP is deployed, that this support extends throughout all of the VoIP components. Are all WAN voice routers, core switches, distribution or closet switches attached to that UPS so that power is continuously supplied to the phone sets on the desktops? Is the UPS sized appropriately to provide for continuous VoIP operation for the period of time until power is either restored or the backup generators are started? Is the VoIP network, in fact, even connected to the ‘emergency power’ circuits? I have seen implementations where they were not. The assumption was that the data network must be connected, but alas it was not as high a priority as other more critical services. So after about twenty minutes the IP phones ceased to operate leaving the organization without voice capabilities except for personal or company issued cell phones.
Even if you utilize an external VoIP service you need to be concerned with powering the phones and other electronic equipment, which connects you to that service. If you utilize a commercial cable provider for your VoIP service make sure that their availability objectives meet your requirements for service. Remember, there is still a box somewhere on your premises where the cable connects to bring you VoIP phone service and that is most probably powered by your commercial power utility source.
Hardware redundancy
Another area of concern for contingency planners who need to maintain high levels of availability for phone operations is the reliability of critical components of the VoIP operations. In order to ensure that the phone is there when you need it there should be sufficient redundancy built into the network and systems to accommodate any component failure without losing voice services. For organizations deploying their own VoIP infrastructure this means redundant switches (core and distribution), redundant call managers or other devices which provide the call processing and routing functions, gateways, routers, etc.
Again if you are utilizing an external vendor or carrier for VoIP services make sure that their recovery and availability service levels meet your requirements if you rely heavily on voice services to accomplish your organization’s mission.
Circuit or link redundancy and diversity
In addition to hardware redundancy it is also a requirement for high availability VoIP operations to have circuit or link redundancy or diversity. All WAN circuits or circuits that connect your organization to the public telephone network, or to other locations within your organization’s network, need to be not only redundant but also diverse. Redundant simply means that you have two or more circuits so that in case one fails the other can take over. Diversity implies that there are two or more circuits or links and that the routing of the circuit is physically different for all of the circuits or links employed. This protects against such things as bundled circuits all being cut at the same time. Perhaps even obtain the circuits or links from different sources. This redundancy is also true of internal links. This means that all core and distribution switches have multiple links to each other to insure continuation of service in case one fails.
In the case of externally supplied VoIP services, redundant and diverse circuits or links should also be considered if you really need to have phone capabilities most of the time.
Sizing and capacity
We have discussed the idea of hardware and circuit and link redundancy. And I am sure that this was all included into the architecture by the designers. However, one of the areas that I find most often overlooked is that of sizing. If we need to fail-over, whether to internal facilities or to other external geographically diverse locations (hot or warm sites), the overall capacity or size of the fail-over circuits or links must be reviewed for adequacy.
When designing in the fail-over capabilities one needs to consider the existing traffic on that link and the increased traffic due to fail-over. If this is not done the bandwidth of the link could become swamped causing congestion, which is a bad thing for VoIP. It could cause conversations to have gaps or distortion and affects the overall quality of the call(s). In addition, if a company, for example, were to reroute traffic from its NYC office through its Boston operations they would need to consider the potential of added costs because every local call could quite possibly become a toll call. Fail-over routing needs to be considered.
Once again if your organization is utilizing a VoIP service provider or carrier you need to make sure that any fail-over sites that will be involved with taking over the voice traffic from one location to another have circuits and capacity for providing the service which are adequate to the task. You could be doubling the amount of calls that your Boston office would be taking if NYC went off-line. Will the circuit capacity or bandwidth accommodate that additional amount of traffic? Can Boston’s personnel handle the overflow traffic? Better find out now before a crisis occurs.
Call dropping
When a VoIP network fails-over it is possible that, if not designed correctly, active calls could be dropped. This should be considered during the design and requirements phase to determine what the requirements, budget, and needs are regarding call dropping.
Final thoughts
VoIP is a technology that offers tremendous opportunities for small, medium, and large businesses. However, like any technology due diligence and due care are necessary for the adopters of this technology. Only organizations that adequately understand the differences between VoIP networks and services and PBX – public switched telephone networks and services will be properly positioned to realize the rewards of this new technology’s use even when adverse situations arise. Think about the design and mitigation strategies that I have discussed above and make sure they are adequate to the task.
Dr. Jim Kennedy is the Business Continuity Services Practice Lead and a Consulting Member of Technical Staff for Lucent Technologies. Dr. Kennedy has over 25 years experience in the business continuity and disaster recovery fields and holds numerous Master level certifications in network engineering, information security and business continuity. He has developed more than 30 recovery plans, planned or participated in more than 100 business continuity and disaster recovery tests, helped to coordinate three actual recovery operations, authored many technical articles on business continuity and disaster recovery and is a contributing author for two books, the ‘Blackbook of Corporate Security’ and ‘Disaster Recovery Planning: An Introduction’. jtkennedy@lucent.com
•Date: 12th May 2006 • Region: US/World • Type: Article •Topic: Telecoms continuity
Rate this article or make a comment - click here |
