When many Canadians think of disasters they look at the horrific events that occur in other countries. But Canada is also vulnerable to damage and destruction caused by nature and by people.
According to the Canadian Disaster Database, the following disasters occurred in a span of just three years:
- Meteorological – Hydrological: Winter Storm
- Meteorological – Hydrological: Flood
- Meteorological – Hydrological: Wildfire
- Meteorological – Hydrological: Tornado
- Biological: Epidemic
- Hazardous Chemicals: Leak / Spill Release
And according to Natural Resources Canada, British Columbia, the St. Lawrence Valley and portions of New Brunswick are also vulnerable to earthquakes. Coastal B.C. notably faces the risk of volcanic eruptions, including from neighbouring Alaska and Washington State, and tsunamis.
Then there are the more common events that disrupt lives and businesses. Like when a truck knocks down a power pole. Or when a contractor slices an underground line.
Finally, there are the even more widespread technology-related disasters. After all, hard drives fail and servers crash. There could also be external issues like Distributed Denial of Service (DDoS) attacks or even ransomware.
Customers rightfully expect to be connected to individuals who can help them when they dial in to order products they are seeing on infomercials, in response to direct mail pieces or flyers, to find out order status or if they have problems with their purchases or billing issues. In turn businesses depend on customers’ orders and their loyalty.
That is why businesses need contact centre redundancy i.e. the ability to ensure that the applications keep running in an event when there a planned or unexpected outage occurs. Customers may or may not know that an outage has occurred, and they may likely be sympathetic when it does. But ultimately, they expect businesses to provide them with the sales and service they expect.
When looking at redundancy, you must plan for the worst and expect the unexpected. The first thing you want to look at is separating the primary and the redundant servers. The further your redundant contact centre server is from your primary server the more resilient your centre becomes. However, physical distance is not as important as ensuring the servers are located on different power grids. Therefore, if the power goes out in your primary location, the redundant location will not be affected by the local or regional outage.
Secondly you want to ensure that the connectivity to the locations are serviced by separate carriers as well as with disparate media types. By doing this, you are protecting against a specific carrier’s outages as well as physical issues, like cut fiber or copper.
Wide area network connectivity
With redundant contact centres, distributed over geographic areas, the next piece of the puzzle is the connectivity between the primary and redundant locations. The connections between these locations should have low latency, either Multiprotocol Protocol Label Switching (MPLS) or with a Software Defined Wide Area Network (SD-WAN) solution.
Most redundant contact centres operate in active/active or in active/standby paradigms. At some point the two applications will speak with each other and their conversations or communication is what transfers information and data back and forth. This continual synchronization keeps both participants up to date with the most current routing information and data configurations prior to an unexpected event occurring.
When an event occurs that the primary server goes offline, customers and agents are registered and routed through the redundant location. If the connectivity is suspect or not up to par, then it will affect the customer experience and the agents’ abilities to serve customers.
When you have primary and redundant contact centres located across geographic areas, the next item to address is the potential failure of trunks into each location. This can be a result of a carrier outage or an outage of your primary location.
One way to respond to this failure type is by having multiple carriers serve each location. Therefore, if you have three Primary Rate Interfaces (PRI) at a location, use two or three different carriers to deliver those services. That way, if one carrier goes down, then the others can pick up the slack.
Another option is to use different trunk types delivering the service at each location. Instead of three PRIs you implement two PRIs each with a specific carrier and the third delivered via session initiation protocol (SIP) trunks. This method would address carrier specific outages and trunk redundancy locally but not enterprise wide.
When delivering any trunk service to the contact centre location you need to work with the carrier to ensure that you have set up call forwarding to your redundant trunks. Therefore, if the service ends at the primary location the calls at the carrier level with forward to the redundant location.
Connecting home based/remote agents
Using home/remote agents is a proven customer contact strategy: including when disasters strike bricks-and-mortar centres. But what happens to them when redundancy takes over? The short answer is, it really depends on how and where the remote users connect to the office network.
Let’s assume the primary contact centre is at the head office and the remote agents are connecting back to it. And let’s assume the head office is no longer functioning and the contact centre had switched over to the redundant location. How do you connect the remote agents?
The best solution lies in having redundant virtual private network (VPN) appliances: one at the head office and the other at the redundant location. The remote agents will most likely re-register to the VPN and reconnect, resulting in a very small (minute or two) loss of connectivity.
Active failover testing
The key to maintaining a redundant contact centre is to ensure that it works. If you set up everything in the most fault tolerant, resilient configuration but you never actively test the configuration, then you still do not have a redundant centre. At this point, you are waiting with fingers crossed that you never need one.
The successful redundant contact centre are those who regularly force their principal contact centre servers into redundancy mode. If you can control when your contact centre server is in redundancy mode, then you’ll be able to identify potential issues with the failover as well as confirm that the technology works as expected.
Protecting the cloud
Going to the cloud for contact centre applications enables companies to have, in most cases, the redundancy they need through the hosting manufacturers or third-party providers. What this means for your organization is less burden on purchasing and supporting on-site hardware because all of it is in the cloud and the only requirement for onsite hardware could be the physical phones. If your company decides to run on softphones clients or WebRTC softphones, then the need for physical phones is no longer required. But should you go to the cloud you must exercise due diligence to ensure that the provider has the tested resilient methods, like those discussed above, in place to ensure continuous service to your customers.
When implementing a redundant contact centre on-premise or from the cloud, or just reviewing your existing redundancy design, ask the question, “What do we do in the event that there is … (choose one of the disaster scenarios from earlier)?” Hopefully the design addresses how your operations will continue during those disaster scenarios. If your redundancy design or that of your cloud application provider does not address them then ask: “Why not?”
Ken Zrobok is director of operations, LANtellligence Inc. He has worked extensively in the LANtelligence Contact Center and on advanced integrations projects. Ken’s primary responsibility and goal has been to interact with customers, defining their requirements and then designing and building complex solutions to meet those requirements.