|
The last fifty years have seen IBM emerging as one of the top vendors of data storage systems worldwide. The pioneer of the world's first storage device the RAMAC 305, has used the headstart to fuel more innovations and is now manufactruing systems that can store 38,000,000 times the capacity of its first product.
Shailesh Agarwal, country manager - storage, IBM Systems and Technology Group, shares with Sigi Achappa of CIOL, the storage division's growth trajectory and talks about some of the promising data storage technologies of the future.
 What are the latest developments on the storage front in IBM?
We have set up the STG Labs, which is into R&D and is focusing on developing new technologies. The lab is like a virtual product center that is being used to develop products that are used not only in India but world wide. Then there is a research lab that is set up at IIT- Delhi.
What are the factors that are driving storage innovation?
From the technology perspective it is the same across the world, there are multiple technologies - the present focus is not so much on technology but on innovation. One of the things in storage more density is being stored on devices the size of a matchbox. There are new parameters which are coming up --for every dollar of IT puchase how much space does it occupy --making physical density smaller has a direct impact per se. Some of these these technologies such as Infrastructure simplification -as storage will keep growing it will become more and more difficult to manage both physically and systems- wise--it has to be made easier. There are three areas we are focusing on:
Consolidation - which involves eliminating components, virtualisation - don't eliminate, but hide it and Unification - where there are multiple different storage systems physically but looks like one system
The requirement for backup consolidation is much higher than server consolidation. There are a whole set of products – consolidation services across geographies such as flash copy, metro mirror over metro distances—which makes a copy of the dayta irrespective of wher the machine is. Infact, Mauritius is becoming a preferred DR location.
Storage management is a major challenge for organizations, how is IBM filling this gap?
As for storage management is concerned, we will never own the information –it is management of the infrastructure. Bharthi, Tata Steel are some of the Indian clients that have outsourced our infrastructure services. There is very high requirement for technical manpower. It will be very difficult for organizations to retain talent. A normal manufacturing company will find it very difficult to hold onto technical talent.
Bharthi has outsourced its entire network infrastructure to IBM. Earlier it was only the larger organizations outsourcing infrastructure management but, now we see smaller companies opening up to it.
At the GSTC, central machines are used when needed – at a fraction of a cost—the first outsourcing model is for DR. Already there are many international customers have India as a DR center.
How do you use the concept of utility computing in the storage arena?
Utility computing gives storage as and when you need it. IT will become a utility, just like how electricity was a utility years ago with captive electricity plants. The impediment to that is bandwidth and mpost of the servers are in the same place. Where people should have started going to data centers – the need has been felt in some countries but it has not yet fully developed. Concerns on sharing, bandwidth are some of the concerns. One of the first examples of utility computing can be associated with the airline industry. Second, the banking industry –transactions can be done on a per transaction type of a model. Utility computing will happen more on the application front than on the system front.
Grid technology is also touted as a promising technology, how will this take shape in the coming years?
Storage started with DAS then moved to NAS the second step in the storage utopia. The 3rd step is grid –not sure as to when it will happen but, a lot of work is going on in the labs. The idea is to use file systems, so wherever the data is I will be able to combine it—we have seen the usage of grid technology in high performance applications. In relational databases it is not easy to do it—maybe other files split and combine easily. The fundamental is the general purpose file system. Then there are talks of a global storage network –it dos not exist anywhere—it is like Internet -- the storage does not lie anywhere. There is a resiliency that is built into the Internet. Here, globally people will come and plug in their storage there—it can be charged based on that.
How long do you think it will take before these technologies actually come into play?
That is difficult to tell. But, there is always somebody who is working on Gen +2. Storage technologies are far from mature.
|
