Definitions of Scale in the Cloud
Last week Limelight Networks sponsored a webinar with GigaOM Pro on the Scalable Cloud. With moderator Derrick Harris, analysts Brett Sheppard and Paul Miller, and our own cloud strategist Jason Thibeault, the panel discussion ranged from defining scale in the cloud, to evaluating use cases, to hybrid clouds, to cloud management and pricing models. Given all of the vague assumptions around cloud services, the ability to hear from experts on real-world considerations was exceedingly valuable. For example, even defining cloud scalability requires a set of assumptions about the audience being addressed. To an IT person, scale means servers, and data centers, and back-up. To a developer, scale means the ability to grow a software application, which means understanding latency issues, computing capacity, and more. Content type also plays a role in scale. Screening the Superbowl poses different challenges than managing an ecommerce site that’s dependent on regular database queries.
And speaking of use cases, one of the big discussion points made during the session was the argument that certain enterprise systems are better positioned for cloud enablement today than others. CRM, HR, finance, and in some cases legal systems work in the cloud because they are, to some extent, detached from a company’s main operations. On the other hand, ERP software and other systems that require a lot of back-end links with a company’s existing infrastructure are more difficult to hand over because of integration issues. In fact, the limitations of enterprise infrastructures are going to continue to exert a push/pull relationship on corporate cloud activity – pushing companies to look for new resources, and pulling companies back because of the IT integration challenges.
