The topic of the "Cloud" was definitely a hot one
back when Bernard Golden wrote a series, named The Case Against Cloud Computing, in regards to the reasons enterprise
architects were reluctant to embrace cloud computing, and it has definitely
escalated since then.
Coming from a background of enterprise software,
particularly ERP systems, and related financial applications, like tax engines
and early payment discounting software, etc, I have been able to observe the
rapid evolution of "the cloud", and thus the evolution of the
barriers to cloud adoption is of interest to me.
In reading Golden’s articles on the topic, we can see that
most architects were very preoccupied with frankly, the basics. For example,
how do we get from point A to point B, how do we migrate the stuff we have On-Prem to the cloud without having to redesign how everything works. Then we had
security, it was argued that cloud providers essentially weren’t competent
enough to handle the client's data they way they could themselves. Golden
also states that one of the main concerns was that Amazon’s core business
wasn’t that of Cloud computing, so folks had a difficult time entrusting their
data to them, in fear all the regulations that applied to their industry
wouldn’t be followed and then also the fact that control would be lost to a
large extent and therefore security would be compromised as a result.
There were other concerns such as the dependency on an
outside vendor and also the fact that compliance measures couldn’t be properly
enforced as a result. In essence though, security, control and “how do we get
there”, seemed to be the biggest areas of concern, as Golden was able to
compile and reflect upon from peers around the industry.
In the spirit of the 10 year challenge that has taken over
social media, these articles were published in 2009; let’s look at how these
pain points hold up today and whether others have surfaced.
Being immersed into this field for the past couple of years
I can say that there have been definite improvements in the area of “how do we
get there”. Most big vendors now have several migration plans that get you from
your On-Prem solution to some sort of cloud offering, be it SaaS or IaaS, and
they have dedicated staff and implementors that work with you to make it
happen. So this is definitely a lot less of a concern, even for big ERP’s like
Oracle EBS, since vendors have come up with “on boarding” strategies and
assessments of your current state and how to get you there. These of course
aren’t perfect, especially if you have super complex existing environments, but
the road is definitely a lot less muddy presently.
In terms of security and the expertise of the vendor, well,
we definitely can say that Cloud computing is one of Amazon’s core competencies
now… and they are not alone, Oracle and Microsoft have very strong offerings
and some of the top talent and best practices in terms of securing your data.
This is especially appealing if you aren’t an “IT” company and your core
competency lies elsewhere.
In terms of Infrastructure as a service, it can be said that
the main challenge I and others have observed is that of performance and
reliability as a factor of cost. Can you get the same service level expectation
and for how much, versus keeping that solution On-Prem. You also definitely
lose a lot of control in terms of when maintenance activities are performed and
how long you lose access to the various systems (more so with a SaaS model). In terms of software as a
service, the biggest pain point is the functionality and losing the ability to
customize that application to fit your business needs. If you are going to a
SaaS solution from a highly customized On-Prem suite of solutions, or solution,
then big changes to your business processes will have to ensue, as you won’t be
able to customize that application(s).
While great strides have been made in the Cloud space, there
is something else that isn’t really a barrier, more so an annoyance, and that
is that many vendors are aggressively migrating their offerings to the cloud
and giving you very little options in that space, in some cases where it would
make sense to have an On-Prem solution or an IaaS offering. Instead, you will
see that the vendor only offers a SaaS flavor of that product you want, and
that is definitely a negative in some regards. If anything truly negative can
be said about the Cloud era, especially in the Enterprise Software space, which
has long lasting effects and probably cannot be mitigated, is that the customer
will lose their ability to choose how they want to play the game.
However, as Larry Hawes notes on his Network of Services
article, it will be beneficial to think of enterprise software as a “network of
functionality that helps all constituents of the business achieve their
objectives” and that architects should think in “networked business terms and
treat software functionality like networked nodes that may be fluently, even
dynamically, related to each other”. I believe cloud computing can and is
helping achieve that network of functionality he describes and helping achieve a more
fluent relationship between systems by many measures, such as standardizing that
“middleware” aspect of data communication and translation by essentially
forcing the use of API’s and removing a lot of the complexity of the past that
could ensue in proprietary environments.
In conclusion, the Cloud had some barriers and pain points
10 years ago which remain today but have evolved into different issues in the
same category. Additionally, new issues and barriers have emerged and need to
be mitigated, but the trend is undeniable and unstoppable in many cases.
References
Golden, B. (2009, January 22). The Case
Against Cloud Computing, Part One. Retrieved January 15, 2019, from https://www.cio.com/article/2431187/cloud-computing/the-case-against-cloud-computing--part-one.html
Golden, B. (2009, January 29). The Case
Against Cloud Computing, Part Two. Retrieved January 15, 2019, from https://www.cio.com/article/2431044/cloud-computing/the-case-against-cloud-computing--part-two.html
Hawes, L. (2012, March 15). Enterprise
Software Architecture: A Network of Services, Not a Layered Stack. Retrieved
from https://www.forbes.com/sites/larryhawes/2012/03/14/enterprise-software-architecture-a-network-of-services-not-a-layered-stack/#5c37fdb07229
No comments:
Post a Comment