Disaster Recovery in a Call Center

Calls centers are the most visible face of BPO today. As most of the call centers are an extension of the companies they serve, the disaster recovery plan for them is different. The data lies at the client’s end and is accessible over high-speed WAN links. Therefore, the biggest disaster that can strike them is to be disconnected from its clients’ sites. The disaster recovery plan for a call center requires building redundancy into the call center, which can be geographical or operation oriented.

Geographical redundancies are location based and can be at the site, state, or country level. For instance, If a company X has a 100 seats call center managed by company Y in Delhi, then its geographical redundancies, also called DR (disaster recovery) centers could be as follows: 

Site level- In addition to the 100 seats, the call center has 25 seats in the same facility reserved for company X which cannot be used for any other purpose.

State level- The call center has 100 seats in one facility in Delhi and 25 seats in another building in the same city/state. 

Country Level- The call center has 25 seats in a different state and city.

Operation-oriented redundancies include the maintenance of these DR centers. These can be called cold DRs, warm DRs, and hot DRs, based on how critical the service offered by the call center is. 

Cold DRs: You have offices in Delhi and Mumbai servicing various clients. Both locations have extra furniture and structured cabling. In case operations are affected in Delhi office, computers can be put up in Mumbai office and the WAN link redirected. This setup would take a few days. 

Warm DRs: Here infrastructure includes computers also. In case of disaster, IT team installs the required applications on these machines. The WAN link is then redirected and executives moved to the new location. This setup would take about 24-hours to complete.

Hot DRs: It is online and ready for executives to move in, it’s also the most expensive. It can be used in conjunction with a country level redundancy. For instance, company X may have 50 execs in Mumbai and 50 execs in Delhi, with provisions to accommodate 25 more in each facility. So when the Mumbai facility goes down, the execs can be flown to Delhi and in the mean time the whole traffic can be directed to Delhi. 

A typical call center setup



Assume a call center called ‘WeServe’ with facilities in Mumbai and Delhi and clients ‘X inc’ in Los Angeles and ‘Y Inc’ in New York. The call center would have a co-location facility in the US (the country where the client is located) where connections from all clients would terminate. This facility would have WAN switches for the companies it’s catering to, which are connected using high-speed fibre links to an ISP in India. The Indian ISP also houses WAN switches controlled remotely by the call center. Now consider the DR example from above, if the call center in Mumbai goes down, then the network administrator can remotely switch all traffic to the Delhi call center. 

Geetaj Channana