Application portability is improving as emerging
technologies enable more efficient migration between
vendors. But data portability is still a major challenge
for many organizations, and data gravity continues
to hamper full migration to the cloud.
Public cloud adoption and migration have become mainstream for many organizations.
It is no longer a question of whether to move to the public cloud, but rather what to
move, when and the manner in which to move it. Issues such as security and perfor-
mance have largely been addressed in the minds of most IT leaders, but concerns about
data gravity and its effects still surface during any discussion of cloud portability.
The concept of vendor lock-in has been present in the information technology space
almost since the beginning of modern commercial computing. For most computer sys-
tem manufacturers, proprietary solutions and technologies were a primary means of
retaining customers, and the term “switching costs” came into being as a way of mea-
suring the impact of moving from one vendor to another.
In the public cloud world, these switching costs have become less about the application
stack and more about the data itself. Just as the gravity of data within an on-premises
environment can prevent movement to even a single public cloud, that same data grav-
ity can often interfere with the movement of application workloads between clouds.
These days, it is relatively straightforward to move an application between public clouds
given the pace at which public cloud vendors are maintaining parity in their execution
platforms. But moving terabytes, petabytes and sometimes even exabytes of data from
one cloud provider to another can prove costly, time consuming and quite risky to
operations.
The Evolution of Cloud Migration and Portability
The On-Premises Data Center
Consider the following scenario of an organization with an on-premises data center
serving users via a web-based application over the public Internet. This common pat-
tern places the application front end, business logic and data back end in a location
entirely under the control of the organization.
SPRING 2019 | THE DOPPLER | 59