It was many years ago, shortly after arriving in the U.S.,
that I moved to Boston and began my career in IT. If it wasn’t for the
generosity of those prepared to give me a chance, I probably wouldn’t have had
any opportunity to get a toehold on what was then considered an industry in its
infancy.
Keydata Corporation, the first commercial time-sharing
computer firm, was my very first employer in the US. Most of the processing was
done in batch, with reports generated nightly. The businesses sent Keydata their
daily transactions logs which we processed for them. There was nothing real
time about it, as I recall.
Perhaps the concept of time-sharing and cloud computing
differs by just that one fine point – cloud computing assumes real time
processing, otherwise, conceptually, these are the same approaches to data
processing! When you look back at early service bureau companies supporting
time-sharing, they were just providing a big resource (for computing) with an
archaic interface. Furthermore, they typically catered for just one type of
client – either scientific, commercial or government.
The demise of Keydata’s time-share model came about when
mini-computers entered the market and IT managers brought their applications in-house,
wanting to have full control of the environment, including all the data. With cloud
computing, IT managers can retain control over some parts of the application
and data, while only farming out components to the cloud that they feel comfortable with running
outside the shop!
Technology, like fashion, seems to be cyclical –
centralized, decentralized, skinny jeans, bell bottoms … etc. However, this
time around, I really do not think that the cloud model will disappear; it is a
logical progression, considering ever growing databases and a need for
processing power, that makes sense out
of all the data we are capturing today. Cloud computing offers alternatives to
when it is no longer practical to own
and manage all the resources that are necessary from time to time; especially
when setting up a data center for a department store in support of Black Friday’s
rush or building out the resources for stock exchanges to be ready for a Facebook public offering. Would sure be
nice, but the cost would be astronomical – in come public and private clouds,
where you can burst (who came up with that allegory anyway?), in case of a dire
need.
Where am I going with this? I think that the ability to take
advantage of cloud computing should be an integral part of every platform, and
I am glad maRunga is here to support the cloud computing model on HP NonStop!
Yes, it’s new. Yes, it’s in its infancy in terms of addressing business
problems, and it will be up to cloud evangelists to establish the models that
will later influence everyone else. However,
the evangelists are out there, and it shouldn’t be surprising to anyone who has
been around NonStop to see the NonStop platform becoming involved. The decision to develop maRunga was made because we knew these evangelists are out there and so, once again, I am looking forward to renewing the cycle of where, in my own experiences, someone comes to help me out and gives us the chance!
Comments