Application Aware Networking and the Enterprise
Following up on last week’s blog on application aware networking, it’s worth noting that this is not a new idea and in fact has been part of the development plans for a number of leading platform developers over the years, to varying degrees of success.
What makes it so compelling these days is that virtualization and the cloud are increasingly breaking down the network edge, making it ever more difficult to manage and maintain network performance through traditional infrastructure-centric methods. With network intelligence, plus a whole range of other control and provisioning mechanisms embedded on the application layer, organizations can better meet QoS requirements (and keep tabs on their data), no matter where it goes.
But this is not as easy as it sounds. As tech expert Pete Welcher points out, it will take a fair bit of coding to instruct the application to determine its own bandwidth requirements, connectivity pathways, access profiles and the like in order to effectively navigate the increasingly complex network ecosystem it is likely to encounter. Likewise, it will be very difficult to introduce a macro element to this so that the application can accommodate not only its own needs but the needs of others too. The last thing anyone wants is a jammed-up network because one app is hogging all available resources.
A really interesting take on this whole subject comes from an anonymous blogger called the Borg Queen. She offers detailed knowledge of Cisco’s early attempts at crafting an application aware platform, how it petered out, and how it was reborn as the Application Centric Infrastructure. In the end, though, she prefers Plexxi’s approach of a dynamically updated policy repository that can be used by admins or the network controller to deliver vital data to applications on a regular basis while apps and network components continually update the repository with the latest network conditions.
But application awareness goes much deeper than mere networking optimization. As I mentioned earlier, the network edge is breaking down due to virtualization, SDN and the cloud. So not only are infrastructure-based management and performance capabilities inadequate, but security is as well. As ChannelPro’s Wieland Alge points out, the shift from physical infrastructure to abstract, logical architecture must be accompanied by the establishment of perimeters around applications and data sets. This requires the adoption of application-aware policies that can be maintained regardless of traffic patterns or physical location. It also provides a greater degree of app isolation should one become compromised. At the same time, it applies the same level of protection to both external and internal threats.
Other vital functions also benefit from app awareness. Backup and recovery, for example, can be simplified dramatically if there is enough intelligence built into the system. Vsqueak’s Andrew Barhorst describes a situation in which Microsoft’s Volume Shadow Service (VSS) can be used to manage backups of SQL, Exchange and multiple other programs to eliminate unfinished transactions or incomplete app files during the copying process. This maintains backup in an application-consistent manner, rather than a crash-consistent manner, to avoid things like an Active Directory repair following the restoration of a domain controller. It’s a small thing, but it removes much of the complexity in the restoration process.
Application awareness, then, stretches far beyond the network into just about every nook and cranny of the enterprise data environment. This is necessary, because before too long, those nooks and crannies will be distributed across great distances.
Photo courtesy of Shutterstock.