Tuesday 19 March 2013

CIO and the Data Dilemma

Traditionally, CIOs were considered to be experts in managing "structured data" - data generated by the ERP and CRM Systems and all systems which have a database structure of certain data fields. CIOs know how to deploy analytic tools around this structured data to derive, and extract meaningful business critical information and serve value to the business to be able to take "correct" and "decisive" business decisions. Also, CIOs know how to plan and manage and maintain their IT Infrastructure and Storage to manage this "structured data". 

So, it all seemed like pretty smooth and sound sailing so far!! One could fairly estimate the volume growth in the "structured data" based on some trend analysis of data growth in an enterprise over a period of couple of quarters / years. All one did was based on the expected volume data growth you provisioned / bought / deployed the necessary storage boxes / storage arrays. Overall, it was a cool and peaceful life for the CIOs!!

Then came the proliferation of mobile devices of all kinds and the availability of "rich" content business information and data generated by all sorts of embedded systems. This has literally made the CIOs throw away all the estimates on data growth and the storage infrastructure out of the window. The CIOs started calling data as "unstructured" for the simple reason that this deluge of information could not be classified, categorized and managed using the traditional Relationship Database Management Systems of their times. It called for the CIOs to look at the entire enterprise generation and management of information in a completed new way. 

Newer technologies like - server virtualization, virtualized desktop environment and cloud computing added to this complexity. Now, the CIOs had to take a relook at the storage architecture because - the servers were virtualized and VDI is also gaining acceptance in enterprises but the I/O piece i.e. the data storage devices still are not as robust as required, not intelligent and not as flexible as a CIO may desire. Technology solutions such flash-based drives or solid-state disks are options available, but the CAPEX costs are high in deployment of these options. Even if some CIOs managed to control and plan for the "unstructured" data deluge, there was another problem. Over a period of time, CIOs had ended up deploying storage solutions by multiple vendors and these solutions do not talk to each other. There were and still are issues on interoperability and interconnectivity of  these storage solutions deployed. 

The other problem faced by CIOs at large was that the "backup and archiving" window started becoming longer and more time consuming. So, the solution deployed were - deduplication and compression technologies. So CIOs, deployed technologies around disk-to-disk-to-tape backup solutions, others storage snapshots based solutions. All of these had their fair share of successes and failures.

Another challenge that CIOs had and still have is that, whenever a CIO buys a server, the server comes with its own hard disk drives. These internal server disks never get used as traditionally CIOs have been trained to use an external shared common data storage - the SAN box. 

So, what next? How do CIOs manage this ever growing, complex data dilemma? Does a CIO invest more on newer storage technologies or explore alternatives to explore and make maximum use of the existing storage infrastructure? I believe, as a CIO, one needs to choose the second option - make maximum use of existing storage infrastructure. The strategy could be - 

  • look what you have,
  • analyse and categorize the data
  • move "hot" data on Solid State Disks (SSD) / Flash Drives, if you already have it or purchase when you do a storage refresh / additional purchase decision
  • use the full capacity of existing tiers
  • look at solutions around deduplication

In fact, storage virtualization and consolidation are emerging to lead the race herein and effective to simply storage management, reduce administrative costs and improve cycle times. CIOs could also look at cloud from a data protection - backup and archiving and disaster recovery mechanisms. It is very clear that in the future, the enterprise storage architecture and its deployment is expected to gain paramount importance for ensuring agility, performance and reliability and for the CIOs to be able to respond faster to business needs for new applications, improve service levels and improve the performance of the data center and reduce costs.

Today, CIOs need to accept and understand that the refresh cycles on enterprise storage are going to shorten and there will be a need for taking a relook at the enterprise storage architecture sooner than before.

No comments: