By Gil Haberman
The past decade has been particularly exciting for enterprise IT and global business. Virtualization and more broadly software-defined-everything drove an accelerating pace of innovation across all layers of the technology stack that continues to support global business. For those of us who were already passionate about infrastructure solutions, this trend translates to greater agility, control, and level of service across the enterprise.
However, addressing the needs of organizations’ branch and remote offices seems to remain allusive. The primary reason is that many treat these remote locations as if they were small data centers, trying to place traditional compute, storage, networking, and data protection in these locations.
This has historically been done out of necessity because these branch offices require the same level of IT performance as headquarters and/or there were few alternatives to infrastructure stacks to support IT needs. However, there are fundamental differences between these environments that translate into how each should be handled from an IT administration perspective. While large data centers hold top IT talent working to maximize utilization using the organization’s latest technology, branch and remote offices are high-numbered, fragmented, limited in connectivity, and hard to reach. With little to no IT staff in these locations, ongoing management is challenging and fly-and-fix missions become IT’s worst nightmare. This results in little innovation that can translate into business performance or savings to the bottom line.
Eight primary needs should be considered when thinking of innovating remote offices—centralize applications, simplify management, provision remotely, ensure remote data is safe, minimize remote data, improve DR, lower RPOs, and streamline data protection. These can be grouped into three key objectives for innovation, efficiency, simplifying IT practices to reduce operational costs; resiliency, improving disaster recovery (DR) readiness and reducing branch downtime; and security, mitigating risk associated with data in remote locations.
Simplify IT Practices, Reduce Operational Costs
The first and most intuitive objective is driving greater efficiency. This is traditionally performed in the data center with converged infrastructure, but a new flavor for remote offices—branch converged infrastructure—tackles efficiency in a slightly different way.
These solutions include WAN optimization and virtualization, which paired together offer efficiency at the branch. WAN optimization accelerates application delivery across distances, effectively enabling organizations to centralize applications without impacting end user performance.
Since these applications serve multiple branch offices from the data center, ongoing administration of these workloads is greatly simplified. For local workloads, virtualization is used to administer and maintain virtual machines remotely, reducing the overall burden on IT. Finally, by adopting a standardized branch converged infrastructure approach, organizations not only benefit from a smaller IT footprint in remote locations, but also eliminate the need to fly out to deploy or replace server, storage, and networking gear.
Once the converged appliance is plugged in, provisioning and control is done remotely—without compatibility or integration complications. This simplicity results in reduced operational costs.
Improving DR Readiness and Reducing Branch Downtime
The second objective is improving branch offices resiliency. Remote locations are notorious for having unreliable and inconsistent data protection and DR practices. While data centers have enjoyed improvements in these areas in the past decade by automating processes and reducing recovery times, branch offices were largely left behind. Many organizations still use tape for backup or infrequent data replication that is rarely tested as part of a DR plan.
The challenge organizations face is that there are many remote locations holding local data, creating a wide geographic distribution of “data islands” that are highly redundant in nature. Therefore, an efficient approach to organizing data is to employ a storage delivery—also known as data centralization—approach where all primary data is managed from a single location, the data center. Subsets of the primary data, or “datasets,” are projected to a specific branch location to serve users. From an end user perspective, data is maintained locally to achieve high performance but once updates are done, changes are synced back to the primary location, aka the data center.
Using a storage delivery approach, the data center always holds recent data from all locations, eliminating the need to set up and maintain backup solutions for each remote site separately. Organizations can then tap into enterprise-grade data protection practices in the data center to ensure backups are maintained properly across the enterprise. From a recovery-point-objective, data is often synced back to the data center, so the data loss impact is minimal in case of an office outage.
Using this approach, DR processes are simplified and result in a recovery time objective of minutes instead of days. Organizations facing a remote office outage can simply start the branch virtual machines in the data center instantly or project the data to an appliance in another remote office.
Mitigating Risk Associated with Data in Remote Locations
The third and final objective is security. Traditionally, organizations have focused on ensuring data residing in remote locations is safe. Given that today most remote environments include a mix of centralized and local applications, it has become particularly important to select a security strategy that addresses both local data encryption as well as securing data in-flight. Interestingly, many organizations, particularly those with remote data in uncontrolled or unstable environments, have started actively looking for topologies that reduce the overall data footprint in those sites as a way to mitigate risks of having data in edge locations. Here, the aforementioned storage delivery approach is particularly helpful—data is largely removed from edge locations and centralized to the data center, with only a small subset of the data stored at the branch at any given moment.
Branch Office Improvement
Today, organizations are in a great position to leverage enabling IT strategies such as branch converged infrastructure, storage delivery, virtualization, and WAN optimization to address the unique needs of branch offices. With careful planning and standardization across sites, it is finally practical to bring the level of service and resiliency capabilities developed in the data center to branch offices and remote locations, which translates into IT infrastructure that supports better business performance despite distance and location. SW
Gil Haberman is a group product marketing manager at Riverbed Technology.
Sep2014, DPS Magazine SWM3005