How to Effectively Manage Data Centers?

by February 13, 2017 0 comments

With the increase in the number of data generated CIOs find it hard to manage the content effectively. One quick solution that comes in mind is investing more money to enhance the capabilities of the data center but in reality, it’s only the half solution of the problem and requires a huge investment as well. Data center could be complicated, yet to manage them effectively, you need to make right decisions.

We spoke to Mr. Sanjay Motwani, Regional Director- Asia Pacific, Raritan APAC to understand the complexity and solutions available, according to him, “The volume of data has increased, the concern lies not only in surge in data but lack of infrastructure to handle the load. Most data centers today are operating only at 30-40% of their compute capacity which is less. Besides, there are other aspects too related to the current situation including network infrastructure or bandwidth that exists in the urban areas as physical infrastructure to carry the information up and down and, point of sale or cost terminals.”

With increase in transactions, the cost terminals need to be scaled up and available at various points of purchase. However, it will lead to bandwidth issues either in the physical lines or at the telecom infrastructure. At the backend, when the terminals are coming into play and transactions are happening, the security of the transactions will need to be taken care of. For digital transactions, OTP generation is real time which has to have a very quick response and is a combination of bandwidth and compute capacity.

All these processes need to operate very quickly and assist the end consumer in real time to make cashless transactions easy. The data centers need to quickly scale up their compute capacity and security aspects to aid the same in the wake of demonetization.

Solutions available to upgrade data centers

Mr. Sanjay Motwani, Regional Director- Asia Pacific, Raritan APAC Ashok Pandey | Pcquest

Mr. Sanjay Motwani, Regional Director- Asia Pacific, Raritan APAC

There are solutions available for physical bandwidth in terms of cables, hardware, etc. However, the issue lies with telecom bandwidth in terms of infrastructure which is not available at the time of transactions with 3G or 4G networks and 2G in rural areas. The telecom industry needs to scale up the infrastructure in terms of the bandwidth availability for mobile transactions. Thus, in the first aspect of spectrum the solution lies with the telecom guys to improve and scale up their infrastructure very quickly.

In the second aspect of data centers, the data center managers need to expand their compute capacity to meet the future demands. This can be done in two ways, first, by expanding the physical infrastructure while adding more servers to the rack. The other way is to increase compute capacity per rack which eliminates the real estate costs of physical expansion of a data center and is a better and more cost effective way of optimizing the existing capacity for the long term.

Data center trends

One of the major trend that the industry is going to witness is optimizing the existing infrastructure by making the compute capacity of a rack from X to 3 X. Data center managers need to see the adequate power available in the rack. Generally, 90% of the data centers today operate on single phase PDUs supplying power to the rack. However, going forward the CIO’s are going to look at 3 phase PDU to the rack which will give them nearly double or two and a half times, the increased compute capacity of power in the rack. This will also allow them to increase their capacity without expanding the physical infrastructure.

The other trend which can be foreseen is Environmental management. With increasing compute capacity per rack, more heat will be generated. Precision air conditioning of the racks will play a critical role here and the data center managers need to measure temperatures at per rack level, from top to bottom with a very high density load. Phase II of environmental monitoring will include Cable management and intelligent cabling system to upgrade the data center.

How remotely manage Cloud data?

Cloud data refers to the data residing in the third party location. For remotely accessing the stored data, data center managers need to have an access to the infrastructure without comprising on security. KVM (Keyboard, Video and Mouse access) is the solution for remotely managing the data stored in cloud which was available in the market for more than a decade but was not utilized completely. The issue with KVM is the security aspect, which means while providing access to a service provider at the time of server issues, there is no direct access to the network of the equipment.

Therefore, there is a need for a mechanism through which two networks can be segregated where the other network will come to play for SLA or service management of the equipment and the first network will be only for owner of the equipment managing the application. This can be managed by 3 phase PDUs in a rack which comes with dual Ethernet ports where one port is used with private network and other is used with service network.

Impact of IoT and Big data analytics on data centers

IoT and Big data is going to have huge and exponential impact on data center in the Indian Market. With this the associated thing that come is storage as now we are talking about digital transactions whose records we need to keep for at least 10 years as per the regulations. This will be a snowball effect but the data center managers need to make best use of their existing storage capacity as the storage of these transactions will hit the roof. This data needs to be stored efficiently, secured and have easy access in real time basis. It will only be limited not by the end consumer’s desire but by the adoption of technology and implementation of infrastructure.

No Comments so far

Jump into a conversation

No Comments Yet!

You can be the one to start a conversation.

Your data will be safe!Your e-mail address will not be published. Also other data will not be shared with third person.