Foreword: In the last issue, we talked about the pitfalls of digital transformation in general, this time we focus on the cloud. What is cloud? It is infrastructure and the only way to digital transformation. However, migrating to the cloud is not the goal; promoting business development is the goal. Also, can you stand being fooled by the manufacturer?
As an important part of digital transformation, the cloud has never received more attention today. "Has your company moved to the cloud?" Today, this phrase is as popular as asking "Have you eaten yet?" when meeting acquaintances.
According to IDC's tracking data on the public cloud market, the scale of China's public cloud IaaS market reached US$1.44 billion in 2016, a year-on-year increase of 65%. It is expected that by 2021, this market will reach a size of 5.88 billion US dollars, with a 5-year average growth rate of 32.4%.
In sharp contrast, the traditional data center market is shrinking. Undoubtedly, enterprise cloud migration is a must-do problem. However, how many companies go to the cloud for the sake of going to the cloud? What about business, planning, and underlying support? What about strategy, what about partners?
Negative Case: More Harmful
When going to the cloud becomes an inevitable choice for digital transformation of enterprises, how to go to the cloud has become the core issue. Move all your business directly to the cloud? What I'm telling you is a liar, just block it.
On Christmas Eve 2012, when everyone was waiting for the festivities and entertainment to come. AWS's elastic load balancing service went awry, causing Netflix to shut down, leaving the hordes of customers hoping for a decent Christmas on the streaming service to remember.
On May 9, 2016, the Silicon Valley NA14 instance of Salesforce.com went offline, causing it to lose power for more than 24 hours. The loss of customer data is massive, and the losses are untold. On October 21, 2016, DYN suffered a series of distributed denial-of-service (DDoS) attacks, and its business almost came to a standstill.
Similar examples abound. Going forward may encounter unknown challenges, but you can't go back or stagnate. What should you do? Only with more thorough planning, more detailed task decomposition, and more comprehensive fault tolerance mechanisms, can everything be done better.
First, it's a systems engineering
Therefore, in general, enterprise cloud migration is a systematic project. It is not a simple arithmetic problem, but an application problem. Just like solving an application problem, there are many ways to reach a goal, and the same is true of going to the cloud, there are many ways. You can choose a public cloud, a private cloud, or a hybrid cloud; you can migrate some businesses to the cloud first, and then complete the migration step by step, or you can directly move all your businesses to the cloud...
But please note that the ultimate concern of solving the problem is the result, whether it can be solved or not, and the method used is not the most critical. The same goes for the cloud. You should not care about the cloud itself, but the application and business, because the cloud is just an infrastructure platform that carries business.
Therefore, the most wanted opinion of the author about the cloud migration is that the starting point of the cloud migration must be the application, not the cloud for the cloud migration. Going to the cloud is because the enterprise business has to enter a new development period. For this reason, the underlying supporting platform must make corresponding changes. There are many options for how to change this, but the most important point is that the cloud must be able to undertake both past applications and future business, so as to be able to link the past and the future.
Second: think twice before you act
With this premise, let’s discuss how to get on the cloud. The second point I want to say is that there must be steps and stages to go to the cloud.
From a macro perspective, the cloud migration is divided into three stages, including the pre-planning stage, the implementation stage during the event, and the post-event optimization stage. Micro dismantling each stage can be divided into many steps.
Taking the pre-planning stage as an example, a lot of content needs to be considered and an in-depth investigation needs to be done. For example: How do enterprises view the move to the cloud? It doesn't matter if it's positive or not, especially how management sees it. Please note that this is very important, otherwise it will only add to the troubles in the later period. When you have no money and no one, you will only fail. In addition, the different states of the IT system determine the urgency of going to the cloud. , and will also affect the steps to move to the cloud to some extent.
In an interview, Laura Guio, vice president of IBM Global Information Technology Services, said that Chinese companies: "They will ask me, what is the best industry cloud computing solution, and what benefits can it bring to my business? Do you guarantee the security of my data? Chinese companies hope to give detailed suggestions, such as how much of my data should be kept on the private cloud and what percentage of it should be on the public cloud.”
This is representative. In short, before going to the cloud, the most important thing is to sort out the internal situation of the enterprise from different dimensions, including operations, business, finance and other links. The clearer the sorting, the less trouble during and after the event.
Third, choose a multi-cloud management platform
Finally, it’s time for the actual combat. It’s time to go to the cloud. The more thorny problem is coming. The first step in going to the cloud is to choose a cloud or build a cloud.
Public cloud is a multiple-choice question. AWS, Microsoft, Alibaba Cloud, IBM, Google... are all big companies. Who should I choose? At this time, if the preliminary work is done well, the progress will be smoother. For example, which applications are suitable for public clouds, pull out a list, and then see which public cloud undertakes more similar cases and has more say. By comparison, choose the more suitable one.
The problem now is that there are too many clouds. Many users will not put their business in one cloud vendor. Maybe part of the business is on AWS, some on Ali, and some on Azure. For users, having a multi-cloud management platform is the most important thing now, who knows what will happen to which cloud will happen at a certain point in the future?
Choose a multi-cloud management platform, this is the focus of 2018. Compared with the hybrid cloud, it is more practical and more in line with the practical requirements of users. Moreover, it is best to choose a CMP that has both cloud services and multi-cloud management experience, and has been deeply involved in the industry for many years. After all, it involves products and services from multiple vendors. Without years of immersion and in-depth cultivation, it is impossible to do a good job in CMP at the technical level and process management level.
Be careful, don't put all your business into private cloud or public cloud in the name of laws and regulations, that's not normal. Hybrid cloud is the norm, and it's no longer controversial.
Fourth, choose a reliable partner
After all, this is actually a requirement for partners. From a technical point of view, it is necessary to consider whether it is open, hybrid integrated, intelligent, and whether it supports public cloud, private cloud, and hybrid cloud, etc., and choose cloud according to different business characteristics. At the same time, there are more industry success stories for reference.
This is the fourth point that the author wants to say. It doesn’t matter what you choose. The core is to have technical support and a reliable, reliable, long-term cooperation and global partner. Only when this point is satisfied can it be possible to migrate to the cloud. This matter has been implemented, and the cloud migration has really brought into play its due value.
At the same time, there are several misunderstandings here to clarify. Going to the cloud is not necessarily more cost-effective than traditional IT application methods. Only the rational use of resources can give full play to the intensive and cost advantages of the cloud; open source software and commercial platforms, there is no absolute who is better. Whoever is worse, open source does not necessarily save money. At the same time, if there is no complete R&D system, open source will eventually become a closed commercial software controlled by others.
So again, technical support is very important, giving priority to and accounting for the largest proportion, and others depending on the situation.
Summarize
On the whole, the momentum of cloud adoption in the whole society is fierce, but the development level is still in its infancy, with many problems and challenges, and only a few people really enjoy the benefits of cloud. Going to the cloud is by no means easy, and even with caution and caution, there are still many unavoidable mistakes. So do your best to prepare in advance and make bold trial and error. Cloud represents the future, but no one can completely replicate another successful sample. Only by learning the methodology and choosing a reliable partner can you go further.
Past articles:
Are you trying so hard to transform just to prove failure? —— Report on the Digital Trap Series (1)