Mark Van de Wiel banner_cloud_sky_scrapers_looking_up

The Cloud is Now

Organizations have adopted the cloud with open arms and for good reason, but not without challenges

Organizations large and small are embracing the cloud. Data security concerns that were prevalent in the early days of cloud computing have largely been addressed, and most IT departments in organizations are positioning cloud services as a core part of the IT landscape. On-premises data centers shrink if not close down in favor of cloud computing.

Cloud computing is available in multiple forms “as a Service”. The following are four common types:

  1. Infrastructure: to obtain access to a virtual or physical server hosted by the cloud provider.
    • Examples are EC2 servers in AWS, Azure Virtual Machines, etc.
  2. Platform: to access technology directly as a service.
    • Examples include a database, MapReduce, and other technologies in the Hadoop ecosystems, a streaming pipeline (Kafka, Kinesis, Azure Event hub), etc.
  3. Software: for direct access to an application.
    • Examples include Salesforce and the Office 365 offering from Microsoft.
  4. Function: also referred to as “server-less” computing, for a pay-as-you-go service per function call.
    • Examples include AWS Lambda and Google Cloud Functions.

 The Business Case for Cloud Computing

The business and technical case for cloud computing is strong given the cost structure and flexibility. Cloud computing turns what is traditionally a capital expense into an operational expense. Capital is no longer locked up in hardware resources business case for cloud computingthat are written off over the lifetime of the equipment. Also, an on-premises data center requires infrastructure (network connectivity, cooling, etc.) and services (VPN, management, etc.) that are offered as part of the service when using the cloud. Leveraging the cloud provider’s compute power opens up flexibility to obtain the right amount of resources required to get the job done, independent of the hardware life cycle. There is of course still an incentive for a long-term commitment to cloud resources, but by sharing compute resources and as cloud providers have normalized the unit of compute power, the cloud offers flexibility that is generally not available in an on-premises data center.

Running servers has become a commodity. Mundane but crucial tasks like taking backups and ensuring systems are available, traditionally requiring a system administrator, have been automated by cloud providers. As such it is becoming harder and harder to justify staff to manage servers or virtual machines. For Platform as a Service additional benefits include services like the installation of software updates. The combination of payasyougo, in conjunction with on-demand scalability, offers the option to shift from using a particular service one day to using another one the next day. Or, as the business is exploring the most optimal way to obtain analytic results, there is the option to try out different services to identify which fits best.

Thanks to benefits like these IT departments have fully embraced the cloud, and as on-premises hardware reaches its end of life the cloud is considered as the alternative to obtain compute power.

Cloud Computing =The End of On-Premises Data Centers?

Does the advent of cloud computing imply the end of on-premises computing? No. Compute resources are not available at any level so at least for the time being some of the largest environments continue to run on-premises. Organizations are also still wary of security concerns, and regulations may put restrictions on where data can be stored.

Does the advent of cloud computing imply the end to on-premises computing? No.

 Hybrid Cloud Computing

The term hybrid cloud is used to indicate the co-existence of on-premises and cloud servers and services, as well as the use of multi-cloud integration. Very few if any organization will be using just a single cloud provider. Think of some of the many applications you use regularly, for CRM to Expense Management to HR. Chances are some if not all of these applications are consumed as Software as a Service, directly provided by the application vendor (who themselves may utilize a cloud provider to host their service). Gartner and other analysts expect hybrid cloud to be a fact that organizations will have to deal with in the foreseeable future.

Cloud Computing Challenges

Adopting the cloud introduces a shift in how to provide access to servers and systems. Tasks that are traditionally in the IT department like defining firewall rules may now end up in the hands of business users. How can an organization ensure data security in such a configuration, also when data sets are stored on the cloud providers’ systems?

Another common challenge is data integration. Most cloud providers use the concept of an availability zone to refer to what is traditionally considered a data center, with data transfer taking place across a Wide Area Network (WAN), between on-premises and the cloud, or between cloud availability zones or cloud providers. Compared to the comfort of data transfer in an organization’s own data center relatively high latency and often lower bandwidth lead to data integration challenges. Add additional requirements to secure data transfer and protect on-premises data stores from external access, and suddenly what seemed an easy implementation becomes a more complicated setup.

In my next post, I will focus on data integration challenges for the cloud in more detail.

Adopting a Hybrid Cloud Strategy? We invite you to download our Best Practice Guide to Hybrid Cloud Data Integration.

About Mark

Mark Van de Wiel is the CTO for HVR. He has a strong background in data replication as well as real-time Business Intelligence and analytics.

Test drive
Contact us