Like in all professions and industries, trends that are well-documented in IT as well show the way. The ongoing business and professional expectations, however, seem to be an excellent regulating system. Depending on the deductible outcome, each direction is sometimes retracted, forced to modify it, either completely stalled, but the opposite occurs in the same way as it is often reinforced or accelerated.
The purpose of this small reflection today is to share our thoughts about these two changes. In two consecutive sections, you can read the two statements.
The first results of cloud computing (later cloud) appeared in the early years after the 2000s, and it came in after 2010. At the same time, its literature developed, business and professional analysts sought to see what the next step would be, what new directions would emerge, in what industries it would be overwhelming, how some areas of information technology would be affected, and we could even list many further analyzes.
Investigations were often not very long-term ones, as the paradigm shift itself resulted in many heresies, whose reception was understandably uncertain, and even their longer-term projections. That is why often they only made estimates of 1-2-3 years.
Consequently, these 1-2-3 yearly estimates may change from time to time and change indeed.
The proposed two bundles deal briefly with the Private vs. Public cloud and Multicloud issues.
If we must point out two aspects that reinforced the spread of the Private Cloud, this has been attributed to the handling of security questions and the interpretation of the law in recent years.
From the beginning of the 2010s, everyone started slower, faster way to the cloud, but due to the issues mentioned above, the public cloud was limited to a large part of the users. But they wanted a cloud where they thought these problems didn't exist, and that was the private cloud.
In the public cloud world, security as a question in the United States has long been a top priority among users. The majority were convinced that service providers are not as careful as they are, the service providers have access to the files, even if they are encrypted, not to mention the US authorities who can search among data without any request and permission in the data centers. The latter expectation also annoyed non-US users, especially the Europeans, and felt vulnerable.
Concerning security, the world started to turn slowly, and the skepticism against public clouds has abandoned. Users have accepted the findings that most of all the corporate and institutional staff who used to attempt unduly to get data, as they know what to look for, and understood the shared responsibility of security model, which the delicate levels are in their hands.
In the United States, the government, the state itself, and its institutions, for example, the CIA and the local governments have been in public clouds for years. It's true, some got even dedicated, so-called government cloud, but this is irrelevant to safety concerns about service personnel previous mentioned, but it's a good sign for the private sector.
The tension seemed to be legally dissolvable as well. However, in the United States, this did not come to the fore because the data did not go outside the country, as there are sufficient data centers within the state too. It was primarily a European problem, and it has got particular emphasis and turbulence in Hungary. The solution could not have been different, as the relevant regulations and laws were re-thought by the actors, including the United States Government, the European Union, and the largest service providers. The change began with the revision of previous Safe Harbor contracts, with the creation of the Privacy Shield, the two continent's players offered comforting solutions to each other.
At the same time, the EU has formulated a unified recommendation for Member States to overcome the impossible situation of the users and practically regulates for all players in the sectors except for the state itself and his critical facilities, that one of the significant achievements of today's IT, the public cloud should be usable for them de facto.
One of the most important forward-looking recommendations is that if a service provider and its facilities are certified and accepted by an accredited auditor in Europe, the user of the member country does not have to ask for a repeated audit on the service when one uses it.
So, the problems as mentioned above caused by external, objective circumstances, seem to be solved by the above-described actions, and we can say, there is no obstacle by now.
And you see miracles, and the trends have begun to change slowly. Estimated expected public cloud ratios continued to rise, while the growth rate of private cloud expectations has been declining in recent years.
It is clear that the seemingly perceived problems of the public cloud-consumption, seem to disappear, but something else must be here because the numbers have changed significantly (see McKinsey's article, 2016).
According to several experts and analysts, one of the main reason is the long wait for success stories about creating private clouds.
In the last 1-2-3-4 years, significant, or even gigantic, private cloud projects have started. The private cloud is also "just" a cloud, and regarding the cloud computing features, it is obvious that there is no reason why it should be "less" than the public cloud, especially if, for example, ones have previously doubted the public cloud services level and data security.
To ensure that a cloud platform can meet these requirements by 2018, it is necessary to provide dozens of services. Without describing these services in detail, we need to know that we expect only essential elements of cloud technology.
Today's data center is no longer VPS and not just an IaaS, but PaaS. Today, these services cannot be operated without high-level software solutions effectively, safely in any way.
This is the so-called SDDC (Software Driven DataCentre).
Private clouds should be able to reach the same level. And that is not easy.
I have no hardware specifically designed and manufactured for me, and there are no solutions providing management and diagnostics according to my specification, which provide perfect integration with all the platform elements and their implemented components.
We have neither installed 100% same tools and environments with the system we are currently installing thousand or ten thousand times, nor we have the full documentation and Tips & Tricks.
Many people try to achieve success on the OpenSource platform, where the community can provide support. But the reality is that, in the absence of proper documentation, they often get in an insecure situation, and finally it takes much more time as it was expected if they succeed anyway
The difficulties hardly mentioned here require an even higher level of professional knowledge from implementers, which is not always available.
As a consequence of this thesis, we get back to the statistical trend shift.
Ones should treat these works as a significant individual project, mostly without any previous experience. Therefore, they are running beside enormous risks. Not just one client has been working on the establishment of private clouds for many years, many deadline modifications have expired, and so much money was spent irrelevantly. Often experiments called PoCs are being conducted, about which management doesn't have even clue.
Due to the novelty of the subject, the management of companies is often patient, but there is also a level of time and money that is too much. Especially when the responsible professionals are not even able to tell the exact time and sum until completion.
Hundreds of companies and institutions are contradicted by the fact that they have been hoping to increase business efficiency through the creation of a state-of-the-art private cloud, and it is getting far ahead of them by now.
The prolonged implementations, let's say failures, drastically reduce the private cloud growth expectations. Ones interrupted their implementations, and many of them have deleted projects from their plans finally. Small part remains at the traditional on-prem deployment, but the majority migrate towards a public cloud according to statistics.
The latter statements are not surprising, as we have already mentioned that the initial concern of security is disappearing, and recent legislative changes help for those the most who are eager to go into the public cloud.
Having had some public cloud experience, users see newer and newer benefits.
Due to the split operation and responsibility, users have nothing to do with the environment (building, power, air conditioning, physical security, audits ...) and the lower levels of infrastructure (e.g., hardware).
As a result of high-level automation and user-friendly services, many tasks, implementations, operations could be covered by staff having a modest knowledge, which is way cheaper. In the case of outsourcing even qualified firms do not require significant sum.
Mostly, the project deadline and cost of implementation can be much more accurately estimated, and the cost of usage is fully transparent and up-to-date. These latter considerations are crucial for decision-makers.
There is no doubt that almost 100% of the services are mature, well tested, used in vast numbers that provide users great safety.
It comes to my mind, and I remember when, during the extensive phase of domestic IT business, salespeople used to chase the customer from quarter to quarter to upgrade, for the latest software version. Of course, it needed three times larger and three times more expensive hardware, but some smart customer opted out of several versions correctly.
Today's "private cloud champions" are continually struggling with the latest, unpublished version. Meanwhile, the public cloud customers sit back comfortably and run on mostly mature releases. What a difference! Of course, I'm talking about primary systems and essential services especially, and not about the continually expanding capabilities and newer services.
Public cloud service providers have felt this trend well and want to further reinforce their acceptance with newer and newer services available in VPC (Virtual Private Cloud).
For example, Amazon Web Services (AWS), has announced his PrivateLink service. With this endpoint enabled in AWS VPC, it is possible to access several additional AWS services directly through the AWS private network, without having to go to the public Internet. Thereby AWS decreases the security risk and avoids communication difficulties coming from traffic uncertainty. If you create a leased line connection to a customer's site through AWS DirectConnect, then using these solutions, we can rightly say that we've created a secure, properly-protected extension of the client environment.
And if someone then wants to realize a private cloud, having the described issues without any legal and other (e.g., critical operation) constraints, take it on himself!
Due to the limits, the previous explanatory sketches cannot be a complete analysis. We flash an approach, a thought about the problems.
It will surely go with us!Apply