The best candidates for cloud databases are typically anything that is running in a cluster, said Elad Efraim, CTO and co-founder of eXelate. “It can be Hadoop, an elastic search cluster where we implement real-time analytics, or a massive, distributed NoSQL cluster where we keep a raw data repository available for low-latency transactions at very large scale. Anything that tends to be in a cluster implementation, anything that can run on commodity hardware—anything that’s dynamic, growing and scaling rapidly—would be a great fit.”
Cloud is also a perfect complement to companies that are looking to execute on mobile strategies, said Matteson. “Many companies are playing catch-up with mobile, and any delays that result from the procurement, installation, and configuration of hardware and software are costly and completely unacceptable. The speediest path to market is via the cloud. We’ve seen this in our own customer base where customers are able to build and deploy mobile apps in a matter of weeks.”
Ironically, in the process, cloud applications “are becoming too siloed” as well, said May. Data being added to Evernote may not be the same data found in a company’s Dropbox folders. “Businesses that can offer a single platform for those companies with multiple applications will have a huge advantage in the cloud game, as they’ll have the upper hand in holding all of a company’s key data in one place.” A number of BI vendors have also partnered with SaaS vendors to achieve this capability.
Big data is also a big driver of cloud, said Matteson. “Much of what we consider to be big data was actually born in the cloud,” he pointed out. “Twitter feeds and web analytics data are good examples of this. In looking to ingest this data, there is no specific need to move it on-premises. There is no advantage to having it on-premises from a performance or data movement perspective.”
Cloud doesn’t necessarily reduce the complexity of data management. Ultimately, no matter how digitized enterprises become, there will be a continued requirement for traditional database skills for some time to come. “If you consider the role and responsibility of the database administrator, his or her workload is really not much different than a few years ago,” said Brent Juelich, vice president of application services, CenturyLink Technology Solutions. “True cloud-based database solutions do enable provisioning faster than traditional methods. In our case, we can provision SQL Server or Oracle RAC in minutes, but database management functions are still the same.” In fact, he observes, “the cloud brings more complexity when a database fails or encounters a performance problem. Trying to troubleshoot in the cloud is like an Olympic sport; there are a bunch of highly trained administrator athletes that each have trained for years in a particular sport. Getting them all to play well together is the challenge and is what separates services from provider to provider.”
Bringing It All Together
Ultimately, the four forces of disruption described above are converging, with highly visible impacts on the data services provided to the business. A working example is the ability of field service representatives to interface with CRM systems, as described by Venkat Rangan, CTO of Clari. These applications “relied on timely and accurate data entry, which was often impractical for teams such as field sales,” he pointed out. However, “recent advances in data analysis make it practical to process large volumes of unstructured content and connect multiple data sources. This opens up new possibilities to extract key insights from rep activity data. Field sales reps equipped with smartphones can now enter data on-the-go. The cloud then provides a centralized location for this data to be processed and stored. So, sales reps have everything that they need to accelerate deals right at their fingertips while high-quality data is stored in the CRM.”