PHOENIX – In EMC's sixth floor suite, high above hustle and bustle of SNW, SearchStorage.com sat down with the...
company's chief technology officer Jeff Nick for an exclusive interview. Talking to the press for the first time, Nick discussed EMC's technical direction and how his years at IBM prepared him for the EMC job.At IBM, you designed and architected technologies, grid computing being one of them. At EMC, your job involves more integration of standards and technologies from acquisitions. How are you adapting to this change? As I evolved at IBM, it became less about writing code and more about setting technical strategy. I spent a lot of time integrating technologies both within and across platforms. Those skills are more amplified now as CTO at EMC. I'm more of a catalyst. I'm not as responsible for the execution of technical strategy, but I work closely with our product management groups to make sure vision becomes strategy and strategy becomes product. So even though I don't write patents anymore, my personal satisfaction comes from seeing cohesion being created across technologies. If others are doing the inventing and I'm able to point them in the right direction, I'm very comfortable with that.
What are the biggest challenges you're facing as CTO?One challenge is uniting EMC's growing set of products and technologies that come from different markets. Defining an architecture that can plug products together and make them work seamlessly is one of the exciting challenges that brought me to EMC.
Another challenge is to create a close working relationship between the CTO office and the groups within the company that are delivering the products. We've formed a CTO council where technical leaders from all EMC groups meet regularly and set agendas. I believe in being collaborative and inclusive and not having all the resources taken out of each group and having them report exclusively to the CTO office.A third challenge is changing customer perception of EMC. Many customers only see EMC as the premier storage company, which we are, but we want to educate customers that it's not just about storage, but about information flow. This involves backup, replication, systems management and data tiering, effectively ILM. A big part of my role is to deliver the message of ILM to customers who are only thinking about storage.
How do you keep a consistent, easy-to-use management framework when integrating disparate technologies from acquisitions?Well, you start by getting the products just to talk to one another. When integrating products together in a seamless way, you have to pull out the components and see where there is overlap and redundancy, but also where there is opportunity to carry information between them.
What is EMC's technical direction over the next year?One direction we're going in is moving to common platforms inside disk subsystems. For example, we have the fixed content technology in Centera and file system technologies in Celerra, and there's no reason we can't pull those together so users can get file-level access and content addressable storage (CAS) access across storage subsystems. We're moving to deliver more functionality, so it's not just, 'Oh you want CAS, then there's only this implementation option.' For content management, we'll continue to push Documentum and Legato for data tiering and the correct placement of data. We'll also be applying the technology from SMARTS, which identifies problems in a distributed fabric. What about virtualization? What is EMC's strategy there? We are coming out with our virtualization product, Storage Router, later this year, which we think has the right architectural approach: virtualization in the fabric. What customers are really asking for these days is not virtualization, but heterogeneous virtualization. Other virtualization products tend to homogenize the physical storage behind the virtualization layer and reduce it to the lowest level of functionality. What emerging storage technologies do you think are interesting and have potential? I'm intrigued by technologies that support integration of data across networks. So grid computing, distributed file systems, global namespace all interest me. Giving the user seamless WAN and LAN access is interesting, but it requires intelligent routing and switching to get a consistent view of a distributed network. These are areas that will become increasingly important over the next year.