Answering a query about a DR site, which was not dependent on existing hardware and software, Papneja said, “The solution can be the eNlight virtual network platform. If there are multiple branches in your organization, then they have to be connected with the DR network. Therefore, the primary site will be the DC site while the secondary site would be the DR site. At the DR site, we would have a cluster of systems, about four or five DR boxes, on top of which the virtualization layer would rest and, on top of that, the virtual machines. These would run consuming the minimum resources that are required for data replication. Whenever the system at the DR site receives more load, it will automatically scale up and attach additional CPUs or RAM. Hence, DR can happen automatically.”
An official present at the discussion raised a query of what was to be done, if the DR site had to be used for MIS, Papneja replied, “MIS replication has to be done exactly like in a data center. If you have the database application in the DC, then there will be a replicated application on the DR site as well.”
It is not necessary to have the entire hardware at the DR site. In fact, if a company is using eNlight, then it only needs a few things. “When everything is happening at your main DC, the only thing that you are doing is replicating data. Therefore, the client only pays for the bandwidth as there is no processing happening,” explained Papneja.
Concluding the discussion, Papneja said that the concept of eNlight was that customers paid less for the DR site; about 10% of the cost of the original DC. When disaster strikes, all of the users are diverted to the DR site, which scales up automatically and provides the resources that are required at that point of time.