Cloud computing from a data perspective can overcome some extreme
challenges many organizations have to know where their data
is and to obtain an enterprise view. So just what impact does
Cloud have on data? The wide community feedback to date is that
data management jobs will be easier in some ways and harder
in others at each of these three levels of cloud computing:
IaaS, PaaS and SaaS.
IaaS - a lot of time is saved in managing
the resource demand curve, negating the need for constant rapid
recalculation of hardware requirements. However, the headache
is now transferred to the Network - as network latency and response
time take on new levels of importance. High latency makes tasks
such as copying data slower. In addition, with the technical
details of the underlying architecture abstracted from the user,
there may be unexpected difficulties not encountered with traditional
architecture which is totally visible.
PaaS - previous issues around development
tools and deployment are now easier, but a steep learning curve
still remains. Data modelers will need to focus more on the
semantic layer, to simplify user interaction. Storage security
permissions also have an added element of complexity.
SaaS - at the SaaS level, modeling activities
do not change; being unaffected by computing in the cloud.The
same issues around accurate and secure data definition, contextual
meaning, and naming still apply to ensure data integrity and
integration across the universe. The integration effort to Cloud
requires an understanding of the relationship of data to the
data/object model provided by SaaS; integration being to date
the most challenging part of the cloud. There is certainly scope
for future improvement in the processes around security, connectivity
and data integrity.
In summary, the impact of Cloud on data modeling
depends largely on the ability of the organization to think
differently about data governance amdn management. Redefining
processes and policies to better fit to the Cloud model prior
to entering the cloud computing domain will make transition
much easier. For many organizations, this may mean going right
back and remedying ommissions such as not having documented
metadata, systems of record, data owners/stewards and governance
practices around maintaining single source business glossaries.
It is fair to mention however, that these issues are not unique
to Cloud. Rather, the option of moving data to the cloud or
data modeling to Cloud is a great opportunity for organizations
to reassess their current data management practices.
