P Sridhar Reddy, CtrlS on emerging technologies in the data center space CtrlS claims to be Asia’s largest data center services provider that runs data centers in Hyderabad, Mumbai and Delhi, offering a combined space of 3,70,000 sq ft. In an interview with Ayushman Baruah of InformationWeek, P Sridhar Reddy, Founder & CEO, CtrlS shares his views on top data center challenges and trends emerging in the space
What are some of the top data center challenges you face?
The top data center challenges we face are related to people and their skills.
What are your Power Usage Effectiveness (PUE) and Data Center infrastructure Efficiency (DCiE) figures and how are they relevant?
Currently, CtrlS’ data center in Hyderabad has a PUE of 1.54 and DCiE of 64.93 percent. The one in Mumbai is operational with a PUE of 1.56 and DCiE of 64.10 percent. However, the Mumbai data center is designed for 1.42 PUE with 70.42 percent DCiE. The data center’s initial module is loaded up to 33 percent on its designed capacity. Once load reaches to 75-80 percent of its designed capacity, CtrlS will be achieving its designed PUE of 1.42.
In India, data centers are operating at a PUE of 1.8 to 2.5 and DCiE approximates to 55- 40 percent, which indicates operation at lesser efficiency. CtrlS is making high energy-efficient data centers in India targeting at 1.3 PUE with 77 percent DCiE. These will be operational by end of this year.
Have you embraced virtualization in your data center? What are the challenges with virtualization and how do you address them?
At CtrlS, more than 80 percent of our shared services are supported by virtualization. Some of the major challenges we face with virtualization are:
(i) Sizing v/s performance of the storage array: In a virtual environment, features like thin provision add good value when it comes to CAPEX but virtual environment is much more IO intensive. When the VM count increases on a shared storage, the latency towards the read/write on to the storage increases, and this has to be handled in an effective manner for better performance.
(ii) Backing up and recovering your virtual machines and their data: As effective as virtualization is for maximizing server utilization, it can create problems for data protection. The proliferation of servers can make backup configuration more time-consuming, increase storage requirements, and complicate backups and restores. Support for advanced technologies such as off-host backup or block-level incremental backup becomes critical to overcome the performance and bandwidth constraints associated with virtual environments.
Have you moved to a software-defined data center (SDDC)? Can you brief us about the trends that you see in the data center space?
No, though we have been evaluating the software-defined network (SDN). We are very excited about the possibilities of these technologies, where threshold-defined actions are possible. The trend that we see is the shift from the existing product level bundling of the SDN to the vendor-agnostic open flow-based panel that is integrated with the cloud panel to provide auto-remediation platform at the PaaS level. This will do away with the mundane tasks of housekeeping and workarounds in a self-healing cloud setup.
Any other innovation you have done in your data centers that you would like to share?
We have been experimenting with various tools and technologies, including the PaaS platform and the Workplace Productivity index. The combination will allow us to define the Information Assurance in terms of systems availability and enabling parameters that directly link the productivity per skill set to the investments made and will be a yardstick to measure the contribution of IT in overall productivity.
Incorporated in 2007, CtrlS offerings comprise of data center infrastructure, disaster recovery, storage, application hosting, backup, linux VPS, hardware, Cloud Hosting, OS layers, managed VPS, network and security solutions.