Rawpixel - Fotolia
Nemertes Research CEO and founder Johna Till Johnson has some thoughts about new Cisco CEO Chuck Robbins and his plans for the company once he takes control of the networking vendor next month. Johnson, writing about the presentation Robbins gave with current CEO John Chambers at Cisco Live, said she was struck by two notions. She first said she was impressed with Robbins' decision to prioritize projects and strategies once he takes the reins. She likes this idea because she thinks Cisco often tries to do a lot of projects all at once. Being able to prioritize could be beneficial to customers and partners. But she is less enthused about Robbins' idea to create a single vendor architecture for network infrastructure, IP telephony and security. Johnson thinks that multi-vendor platforms are what customers are looking for right now, not a one-stop shop.
Read more of Johnson's take on Chuck Robbins' vision for leading Cisco when he becomes CEO.
Hadoop losing steam as analytics engine?
There is increasing evidence that Hadoop is no longer the go-to big data engine of choice for enterprise customers. New York Times writer Quentin Hardy says that Spark, a faster data processing engine created by Databricks, may become the new choice for many companies. Why? Mostly because Spark can analyze data from phones, sensors and Web browsers in a matter of seconds, while Hadoop can take minutes or hours to analyze data. Hardy says Cloudera, the industry's largest Hadoop company, is already making the transition to Spark. The advent of a new way to crunch big data is just the beginning of bigger changes that may affect the big data industry, Hardy says. HP, for example, is developing a computer architecture that relies on memory rather than processing.
Read more about the potential impact on the enterprise now that Hadoop is no longer the number one resource for data analytics.
Group-based policy: an open source framework for policy-based networking
Enterprise Strategy Group analyst Dan Conde says that if Group-based policy (GBP) wants to extend its reach, its backers need to keep accepting contributions from a variety of vendors. GBP is an open source framework within OpenStack and OpenDaylight that controls networks through high-level abstractions. Conde says that GBP has the potential to become a foundational framework for policy-based networking, spanning compute, security and storage.
Conde explains three key elements of this project:
- Even though the origins of the project come from Cisco's Application Centric Infrastructure initiative, contributions from third-party vendors have enabled it to become independent and relevant to the entire networking industry.
- It supports service chaining, which allows it to be used for network functions virtualization.
- In order to become accepted in the open source community, it still needs to make changes to its user interface.
Read more about how Conde says Group-based policy can become an independent open source framework in the networking industry.
Should Cortana be retooled for enterprise employees?
Current Analysis analyst Brad Shimmin wonders what the possibilities would be if Microsoft took its personal mobile assistant, Cortana, and retooled it as something that would serve as a personal assistant to enterprise employees. Cortana learns the behavior of cellphone users by keeping track of their browsing, email and social media interactions. It can remind a user when he needs to call his mom on her birthday or make a doctor's appointment. Shimmin says that this kind of technology in an enterprise environment could allow Cortana to become more contextually aware: interacting with coworkers' emails, company databases and more. "What if it could expand its contextual awareness to include not just your inbox, but everyone's inbox, your corporate ERP database, even an entire Hadoop cluster, where your company might store billions of rows of analytical gold?" Shimmin writes.
Microsoft has started the preliminary phases of creating an "assistant" like this, says Shimmin. And it is not the only company to do so. IBM is working on similar technology using its Watson analytics model. Shimmin doesn't know when a virtual assistant like this will become commonplace; it will take data scientists, business owners, security experts and users working together to create the ultimate product.
Read more about what Shimmin expects for the future of the virtualized personal assistant.
SN blogs: Is Spark the next Hadoop?
SN blogs: Data warehousing with cloud and Hadoop