CIOs and IT managers should not build silos in a datacenter. The last thing you want is to have two storage platforms and buy the third one which does not work with the other two, Silos make it expensive to manage storage, says @PhilDavis, #HPE Phil Davis, VP & GM, Storage, HPE Asia Pacific and Japan, spoke about the trends in storage in the APAC region, adoption of flash storage in India and how HP is faring in the market. He also explained the factors that CIOs should consider while implementing flash storage. What are the trends in storage in the APAC region? Is the flash storage market picking up in India? And why? The move towards digitization is the big trend. As one wants to access information quickly, there is lot of stress on the storage systems to operate fast. The next big trend is the rapid move towards flash based storage systems. With business requirements, the flash prices have come down significantly. The growth of spinning disks have become flat. In India, flash adoption is gathering momentum. At present, in APJ Australia and Newzealnd are number one in flash adotion. In Asia, India, Singapore and Hong Kong-- are pretty close in flash adoption. Flash technology is picking up in India as it is more effective. Infact, it is becoming mainstream. At the same time, flash is cheap and delivers high performance, with less than 90 percent of physical storage. It saves money as one is buying less storage to deliver the same performance. Flash is also low on power. That said, it saves 75- 80 percent of the power consumed by spinning disks. It improves organizational productivity, as one can do coding and the next generation of coding in a quicker fashion. Moreover, bugs can go down by 18 percent and there are chances of only 2 percent failure. Also, it helps to do real time analytics to make better business decisions. What do you think will drive flash storage in India? Digital disruption will drive flash storage in India, especially in taxi industries, government, hospitals, public education and manufacturing. Also IoT, where every sensor produces some amount of data that needs to be stored somewhere to do analytics. How is HP doing against competition in the flash storage space? Well, our major competitors bought a startup in flash storage and their new product does not fit into their portfolio. It becomes expensive to operate, as they spend more money operationally. We have taken a strategically different approach. We went to the HP labs and built up Flash systems from the ground which work with other disk based systems. Interestingly, we offer a mixture of flash and spinning discs as some customers who want low cost storage prefer discs. We have product superiority with respect to low cost and data protection. The same thing which can be done on spinning disks can be done on flash systems. Our strategy is to innovate. We will also invest in intelligent capabilities which gives better efficiency with investment. What are the don’ts for CIOs and IT managers when it comes to implementing flash storage? Many customers think flash is more expensive than spinning disks. They should not assume that flash is premier product, as it is quite mainstream. At the same time, we strongly believe that one should not build silos in a data center. The last thing you want is to have two storage platforms and buy the third one which does not work with the other two. So you want a flash system that works with everything you have. Silos make it expensive to manage. The biggest thing is to optimize the application over time to take advantage of performance. How has the advent of cloud impacted HP’s storage business? Well cloud is growing and will continue to grow. However, people are nervous about moving production data on cloud due to security, reliability and performance. Cloud has not disrupted storage. Today, customers want the benefits of cloud. They want to pay as you go. They don’t want to spend $ 5 million dollars on cloud for five years in one go. Many customers don’t want to plan and manage storage. There is a bigger priority on compute.
No comments:
Post a Comment