CIO Specialist Advisory LLP                                                Leveraging excellence and value through collaboration
 
Picture
DD Mishra, Partner, CIO Specialist Advisory LLP, debates seven myths which surrounds the success of cloud computing today. As per Mr Mishra, migration to cloud should be well thought of move based on thorough due diligence conducted with utmost seriousness.  "Large businesses will find combination of public and private cloud more interesting from both business and financial perspective", opines Mr Mishra.

Cloud is on the verge of becoming a mega trend and it is necessary we debate this topic to a great depth and discuss it openly to remove any confusion and look at cloud from a different window of opportunity. I have tried to capture here some of the myths which exist around cloud for a healthy debate in my endeavor to demystify. I am sure there are better opinions and views which are available than what is being presented here and it will be a pleasure to debate them on this forum.

1.  Cloud brings significant cost savings
This is debatable and this may not always be true. There could be situations where a cloud could be more expensive. We often overlook several associated costs to compute the cloud expense. More often this myth is created as rhetoric by product vendors to make a business case for cloud computing. If you are going for an OPEX model with public cloud, your accounting becomes simple but this also comes with an overhead of additional charges. The logic is same as buying a car and renting a car where renting over a long period is more expensive.

If you are having an environment where the resources are largely underutilized with occasional or seasonal peaks, the cloud option could bring cost advantage. The cost benefit analysis should incorporate the bandwidth usage, any license fees, cost of hosting etc vs server cost, maintenance, electricity/fuel, datacenter charges, staff salary and communication/connectivity, network charges etc. which will help us identify true benefits. The holistic benefits of depreciation, amortization, tax advantages and other financial parameters should also be factored in for exact calculation of possible advantage we are expected to get from cloud computing over a period of time and should be compared not in isolation.

2. Cloud will make CIO and existing IT team redundant
This is partially true for cloud. Everything depends on how you implement cloud. Most often for enterprise IT, the hybrid model is adopted as the best bit. This would mean you still may need to manage things internally but yes it may bring some amount of optimization necessary. In addition, this depends on level of cloud adoption happening for example SAAS, PAAS or IAAS. For SAAS, the dependency on IT team is lowest and for IAAS, it is highest. For large IT setups, we are yet to see full SAAS to come and possibly it may not happen in near future. But for SMBs, yes cloud has the capability of reducing the size but this would mean growth of cloud providers who will hire those skill sets which it is replacing elsewhere.  

As regards to CIO, his/her role could be more important than what it is today as he/she will have more power in hand to support business. It only depends whether the CIO is business focused or technology focused. But chances are remote that it will have impact on the role of the CIO as you may still need someone to be managing your IT environment and deal with business demands.

3. If I am done with virtualization, there is not much benefit from cloud
This could be incorrect as cloud will open additional window of opportunity for optimization which is untapped through the virtualization. Most of the cloud computing platforms run on top of the existing virtualized environment to pull together the resources. Virtualization is a technique that allows you to run more than one server on the same hardware box. Cloud computing is not associated with a single box but it can run on several boxes to pool and glue them together to form a service. It has been developed as a concept which allows you to use the infrastructure as a service and extend it further to platform or software as a service.

4.  Cloud implementation is expensive and comes at a cost
One has to be informed about various options. Yes some COTS products can cost a lot but there are cheaper enterprise level options and open source cloud platforms available in the market. Some of the popular names which come are eucalyptus, Nimbus, OpenNebula, OpenQRM, CloudStack, OpenStack etc. Mostly the product vendors at times tend to confuse consumers. Due diligence is necessary before finalization of products to build the right perception about the cost, control, security and efficiency. Hiring a cloud consultant at times is a wiser idea in taking right decision before cloud adoption.

As far as running costs of cloud is concerned, the pay as you go services need a careful evaluation. Amazon public cloud comes with an option of providing EC2 reserve capacity which gives advantage of lower price than “pay as you go” option. Knowing consumption patterns will help in reducing the costs by making combination of reserve instance and pay as you go services in a manner that costs are optimal. In addition one needs to compare apple with apple as just comparing the server costs does not help. One has to compare overall DC, power, infrastructure, network,  fuel, support etc into account when making cost benefit analysis of cloud computing.

5.  Public cloud has significant security concerns
The same dilemma existed during outsourcing and we are discovering the wheel again. Most established cloud providers are using security standards which can be trusted and at times far higher than the consumers. You can check for security and compliance standards of the provider and choose the right provider who is closer to expectations. Not all cloud providers may have similar standards but there could be few who can truly address the level of security and control standards need to be followed.
The success of Salesforce has to some extent busted this myth. Moreover organizations have been sharing the sensitive payroll data with outsourced partners for decades. Outsourcing has allowed partners to manage the internal data as well and this has been there for some time. The resistance due to concerns of security over cloud is expected to fade away in time.

6.  Cloud will cannibalize existing hardware business
I was reading an interesting news on crn.com which states “IBM expects the adoption of cloud computing to add significant growth to its business over the next five years even as it cannibalizes much of its existing business.”  The growth of cloud will lead to growth of overall IT spends and if this continues, there will be more computing power needed. There will be some realignment of smaller business who survived on supporting and repairing the hardware and the traditional sources of IT revenue will undergo some course corrections but at an overall perspective, the cloud will help the growth of business of hardware manufactures as demand may go up.

7.  Anything or everything can be pushed into public cloud
Not everything is a candidate for public cloud immediately. The overdrive to put everything using a big bang approach should be reviewed with great caution. Often we see overdrive to put applications which consumes most resources towards public cloud but that could be counterproductive as we should look at those which consumes least and idle time is high.

In addition highly integrated business critical applications are not the right candidate for cloud immediately. From the practical perspective while small and medium business can have 100 percent public cloud implementation, large business will find combination of public and private cloud more interesting from both business and financial perspective. A thorough due diligence by consultants would be necessary before a decision is made what can move and what cannot move onto the cloud.

Many of these myths have been discussed on different forums and blogs by experts and eventually they will become history as the resistance to change becomes weaker over time.


This article was first published in InformationWeek India by the same author. 

 


Comments




Leave a Reply