Shulou(Shulou.com)06/03 Report--

Cloud computing, as an important supporting technology for digital transformation, can reduce IT complexity, shorten delivery time, make business more agile, improve resource utilization rate and effectively improve IT cost structure. On the basis of cloud computing, it has nurtured the development of digital innovation applications such as big data, artificial intelligence and Internet of Things. Therefore, it is inevitable for enterprises to apply cloud and adopt cloud services to accelerate digital transformation, and it has become the consensus of all walks of life.

Enterprise IT needs to consider multiple factors in the process of adopting cloud services, including time, cost, performance, high availability, cost performance, compliance, security risk, legal compliance, and protection of critical information assets, and it is difficult for a single cloud service provider to meet all the needs of the enterprise at the same time. According to last year's annual Cloud Computing Report by Right Scale, 81% of enterprises surveyed worldwide are using a Multi-Cloud architecture, with 10% using multiple private clouds, 21% using multiple public clouds, and 51% using hybrid clouds. On the whole, the acceptance of cloudy weather in foreign markets is very high. According to the data comparison of last year's ICT survey, the proportion of cloud computing applications in China's enterprises is 58.6%, of which 36.4% use public cloud, 14.1% use private cloud and 8.1% use hybrid cloud. Although this does not directly explain the use of multi-cloud by domestic enterprises, it can be inferred that the proportion of domestic enterprises using multi-cloud ranges from 8.1% to 58.6%.

There is a wise saying in the investment world: "Don't put your eggs in the same basket! "People are warned to prepare for more than one hand, lest the risk comes and they cannot resist it. This idea is also applicable in the field of cloud computing. People all know that relying on a cloud is easy to be "kidnapped", while using multiple clouds can spread risks and be free to do whatever they want. The advantages of adopting "multi-cloud" are also obvious. For example, enterprises can select the best services in various fields to form the best combination to meet business needs; global resources of multiple cloud service providers can be flexibly scheduled to help global business expand rapidly. So what business scenarios are suitable for multi-cloud, and what are the risks and challenges of multi-cloud for enterprises?

First of all, let's talk about multi-cloud business scenarios. From the current needs of large and medium-sized enterprises, there are mainly the following application scenarios:

Cross-regional business deployment: In order to build a national or even global IT support system, large and medium-sized enterprises cannot cover all regions with their own private cloud, and the construction cost is too high. They need to rely on multi-cloud capabilities to achieve global deployment. For example, if a large company sells electronic products abroad, it needs to connect the sales system of the enterprise abroad, which is realized through local private cloud and public cloud. Cost-effective public cloud services: enterprises dynamically schedule according to the service types, functional and non-functional factors provided by major public clouds, considering discounts, layout of public cloud vendors, etc., and adopt cost-effective cloud services, which need to be scheduled under cloudy conditions. Adopt public cloud services with advantages: Enterprise private cloud cannot meet all IT requirements, and different public cloud advantage services are selected for different application scenarios. For example, A's service is used for artificial intelligence applications, B's service is used for cloud storage, and so on.

Cloud disaster recovery of critical data: There is a gap between the enterprise's own private cloud operation and backup management capabilities and large public clouds. Consider the cost performance and regularly backup some systems or data to the public cloud for disaster recovery.

5. Global high availability and performance requirements: The private cloud of the enterprise itself cannot cover all regions to provide high-performance services for customers, and the CDN and global Load Balancer capabilities of the public cloud can be utilized to provide better high availability and performance; Short-term business explosion scenarios: For some enterprises, the business has strong peak and valley characteristics, and the enterprise needs to consider the business support and rapid explosion scenarios, as well as the overall IT cost of the enterprise. Consider adopting multi-cloud. Some Internet innovation companies (such as online e-commerce and training companies) need to cope with short-term resource bursts, while releasing resources during business troughs to meet business needs while saving costs.

Next, let's talk about what risks and challenges cloudy poses for businesses. First of all, since there is no standard for cloud service architectures and APIs of different vendors, each cloud service provider has its own unique products and proprietary APIs, so each cloud service provider needs to be adapted when using multiple heterogeneous cloud services. In application integration, API aggregation and transformation are required, which puts forward higher requirements for flexibility of integration. After using multi-cloud, it is impossible to avoid cross-cloud migration of data. In terms of data consistency, low latency and how to efficiently utilize multi-cloud data, realizing unified management of data and safe and protected flow become new risks and challenges.

Faced with key customer demands, cloud developers need to address the challenges of how data flows safely and freely across the cloud and how applications and data are managed across the cloud. In order to use multi-cloud well and complete digital transformation, the key point is to establish a multi-cloud management system with comprehensive functions, centralized sharing and independent control, support agile deployment of relevant businesses and big data analysis application, and empower enterprise business through data application.

Cloudy clouds also bring more complexity. Containers and orchestration software can make applications portable, but their customizations and data (due to data persistence issues) may not migrate smoothly to another cloud platform; there are also multiple dependencies of applications that need to be considered. Furthermore, data traveling across multiple cloud platforms will be exposed to more risk, at least in theory, as more touchpoints expand the scope of security risks. When data migrates from one cloud platform to multiple clouds, a lot of data can be exposed. In addition, staff issues need to be considered. Enterprises adopting compute services in AWS, Azure, and GCP struggle to deploy enough talent to support security, compliance, and government requirements for each platform. Some companies have gone from adopting cloud to reverting to their original operating environment because it is difficult to keep up with the pace of work. Enterprise IT teams must use tools across all platforms and establish good practices for backup and security. This can be very challenging.

Cloud migration also poses risks to the security of the overall enterprise environment. For example, some public cloud applications have established processes that do not necessarily match the business requirements, which will generate some additional processing time overhead; on the other hand, cloud services are accessed through the Internet, network quality, service bandwidth, concurrent user access, etc. will cause congestion, thus affecting the quality of service of the business. Such risks require enterprises to consider network and business concurrent access in advance, and customize reasonable QoS, SLA and other policies with network service providers and cloud service providers. Of course, for large and medium-sized enterprises, due to the complexity, high coupling degree and wide geographical distribution of IT systems, it is impossible to carry out one-size-fits-all migration in the process of applying multi-cloud, which often needs to be carried out in modules, paths and stages. Moreover, the cloud migration process may involve a large number of application system reconfiguration and replacement, etc. The technical difficulty of cloud migration has also increased under the influence of multi-cloud factors. Therefore, how to minimize the impact on business and risk in the process of accelerating cloud, and realize the "smooth cloud" of enterprise applications has become a major problem for large and medium-sized enterprises.

Today, the trend of multi-cloud is obvious and widely used. However, multi-cloud management is still a complex and realistic problem. Cloud management platform is the inevitable choice to realize multi-cloud management.

The definition of cloud management platform first comes from Gartner, an international authoritative research organization,"Cloud Management Platform (CMP) is a product that provides unified integrated management of public cloud, private cloud and hybrid cloud. Key cloud management platform capabilities include unified management and scheduling of hybrid cloud, multi-cloud environments, system imaging, metered billing, and workload optimization through established policies. More advanced products can also integrate with external enterprise management systems, including service catalogs, support provisioning of storage and network resources, allow enhanced resource management through service governance, and provide advanced monitoring to improve performance and availability. "

The cloud management platform can specifically meet the following management needs: 1. Cloud management platform realizes unified management of multi-cloud 2. Cloud management platform to achieve cross-cloud resource scheduling and scheduling needs 3. Cloud management platform to achieve multi-cloud governance 4. Cloud management platform realizes unified monitoring and unified operation and maintenance of multiple clouds 5. Cloud management platform enables unified cost analysis and optimization of multi-cloud 6. Cloud management platforms help developers build apps across clouds based on APIs for unified delivery.

The development of digitalization has gradually popularized the multi-cloud model, and MSP, which represents multi-cloud management services, has gradually become well known to the public. MSP is a further extension and development of CMP Cloud Management Platform, a neutral multi-cloud management platform and a cloud management service provider. According to the definition given in the series of standards of Cloud Management Service Provider Capability Requirements issued by China Academy of Information and Communication, cloud MSP must include three parts of capabilities, namely "consultation, migration and management." Consulting capabilities include enterprise basic capabilities, system evaluation capabilities and planning and design capabilities; migration capabilities include migration plan capabilities, migration implementation capabilities and migration verification capabilities; management capabilities include multi-cloud access capabilities, resource management capabilities, operation and maintenance management capabilities, operation management capabilities and security management capabilities. At present, cloud MSPs that can meet the above conditions are mostly data center enterprises. Because of their strong resources and customer service capabilities, these enterprises can better meet the complex needs of users for multi-cloud management. The value of multi-cloud management lies in the fact that it can deeply understand the business and IT environment of the customer enterprise from the early stage of cloud, so as to predict future problems and provide full-stack services and customer experience. Full-stack cloud can meet the resource requirements of a large number of application development delivery through self-adaptation of multi-cloud resources for applications. Full-stack cloud provides enterprise-level consistent multi-cloud infrastructure, helping enterprise users build innovation platforms, unlock data potential, optimize platformization effects, build total factor productivity, and enhance global market competitiveness.

In the digital age, multi-cloud is an IT strategy, and IT is like an accordion. At any time, one part is expanding and the other part is shrinking. The shrinking part is standardization, where a variety of options are gradually reduced to a standard; the expanding part is innovation, such as machine learning, chatbots, RPA, artificial intelligence, and companies need to be able to choose and try out countless options. Therefore, multi-cloud management is a process of continuous evolution. The standardized place is standardized, and the innovative place is innovated. Any current giant cannot guarantee that it will always be invincible in the future. It needs to keep pace with the times and continuously iterate forward to cope with the unknown and uncertain future!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.