Managing remote retail stores and remote bank, insurance, and healthcare branches, consistently challenges IT, because these remote outlets require the same quality of service (QoS) and access to networks and data as headquarters, even though they are geographically distant from central IT infrastructure. In recent years, IT has attempted to attack this problem with a two-pronged strategy: placing remote servers in branch and retail outlets that are capable of running the business by storing and later forwarding transactions whenever corporate resources are unavailable; and tunneling remote sites over a virtual private network (VPN) to corporate systems for “live” processing with central IT infrastructure in a secure environment.
Now, however, cloud resources present a third alternative for remote office processing that can potentially simplify network management for IT in some areas (while it adds complexity in others). In fact, cloud can accelerate the provisioning of system resources for remote sites; reduce IT hardware and support costs; and enhance connectivity, availability, and failover possibilities in the event of disaster.
A cloud-based solution can also build in redundancy of systems and access for purposes of failover at the same time that it gives remote office users access to a shared pool of resources that can potentially be paid for on an on-demand basis. Many cloud services providers now feature secure, Layer 3 tunnels into their resources, which isolates VPN traffic and ensures that IP packets are placed on the correct VPN. They also offer Triple DES (3DES) or bank ATM-strength data encryption. This combination of throughput, access and resource security allows remote offices to exist totally in the cloud. It also eliminates the need for these offices to run through secure corporate VPNs, where bandwidth can be more constricted.
Cloud services providers can also implement network business rules from corporate IT, including firewall rules that address timed access, where access to resources is permitted or disallowed at specific times of the day (if desired), for a specified group of servers in the cloud.
As with the corporate VPN, data traffic and loads can also be prioritized in the cloud to ensure that quality of service exists for the most mission-critical processes.
Cloud services providers also provide performance and resource utilization reports, along with monitoring services that can offload support for remote offices from corporate IT, potentially reducing operating expenses.
Given this set of capabilities, it makes sense for IT to at least consider a cloud option for remote office support. The planning process should raise and consider the following questions:
How do we seed the cloud?
Initially, remote office data and applications have to be loaded and tested on the cloud, complete with the inclusion of corporate business rules that apply to both. Most of these data and application assets are too large to be transferred over the network in a timely manner — so it might well be the case that traditional (disk and tape media) means are used to do the initial load.
How do we keep data and applications synchronized?
This is the conundrum of the cloud. While corporate IT shops for management software that gives end-to-end visibility of application and data processes within its four walls, even the best management software is going to stop at the cloud. If there is a problem in the cloud, it's likely to be more generic in nature. All corporate IT will know is that it is up to the cloud services provider to solve the problem. The other element of this is keeping data synchronized, so that if remote offices run off the cloud, there at some point is a process to import and synchronize that data with data residing in the central corporate data repository.
Do we have failover and business continuity?
The cloud (and the corporate data center) may have the potential to failover for each other in event of disaster. This should strongly be considered as an option, and if it is viable, it should be incorporated in the corporate disaster recovery plan, and then tested with the cloud services provider. Additionally, there should be failover strategies in place for the cloud to failover its own resources (given a failure within the cloud). All of this ensures connectivity for remote offices.
What if we have to de-convert?
Business realities change and today’s solutions don’t always carry over into tomorrow. Any strategy employing an outside cloud services provider should also have a built-in path to corporate reconsolidation of resources if for some reason the relationship with the cloud services provider terminates. This transition to either the corporate data center or another cloud provider should be well orchestrated and transparent to the remote offices that depend on continuity of data and applications for their businesses.
For more information, see:
Should Starbucks Be Part of Your Business Continuity Plan?
Considerations for an Efficient Storage Infrastructure
Faster Implementation: The Real Benefit of Cloud Computing?
        | 
 |
