Machine Learning Techniques and Analytics for Cloud Security. Группа авторов

of public cloud, it has developed much from its early date but still it is not 100% secure. There are always threats of data breach. Migration of sensitive need special care as sight alteration in business sensitive data might cause severe problem in the business.

       How to integrate public cloud with existing on-premise system: Amalgamation of public cloud onto an existing on-premise system often needs several alterations in the working of the existing on-premise system. Overall performance of the system should always improve with the addition of the public cloud, and it should not degrade.

       How to manage common backup and disaster recovery: Data need to be backed up to ensure reliability and availability. Backing up of all the data both in private and public cloud need to be done. At the same time, the system should be able to handle catastrophic failure or disaster. How to maintain a common routine for all the operational data to accommodate those situations is key to the success of hybrid cloud deployment.

      Building a hybrid cloud is a complex procedure but successful implementation will provide scalability, flexibility, security, and cost saving. More and more organizations approaching toward hybrid cloud for the current benefit and future growth.

      1.3 Comparison Among Different Hybrid Cloud Providers

The major famous leading cloud computing vendors are Google Cloud, AWS, and Microsoft Azure. They have their some advantages and disadvantages. These three leading cloud providers have important role in the PaaS and IaaS markets. Synergy Research Group reported that the growth of Amazon is very significantly high in overall growth of market. It possesses a share of 33% of cloud market throughout the world. In second position, there is Microsoft. Microsoft is very fast growing and in the last four quarters, and its share has been increased by 3% and it reaches at 18%. Nowadays, cloud computing is become much matured. It is becoming hybrid cloud, and it also becomes more enhanced as market share. New trends have come to improve cloud computing system in 2020 than that of 2017, 2018, and 2019 [17].

      Hybrid cloud [17] provides strategy for enterprises that involve operational part of varieties of job in varieties infrastructure, whether on private cloud and public cloud with a proprietary different layers at the top level. Multi-cloud concept is similar kind of but not to involve any private cloud. Hybrid cloud is the most popular strategy among enterprises; 58% of respondents stated that it is their choice able approach while 10% for a single public cloud provider and 17% for multiple public clouds.

       Microsoft Azure Stack: Microsoft is a popular vendor that provides hybrid cloud. Because it has huge on-premises legacy. The services of MS Azure are developed on Windows Server. The .Net framework and the Visual Studio provide better features of on apps for their smoother services [17].

       AWS Outposts: Amazon’s Amazon Web Services (AWS) is a one of the best product. It is one of the most popular in market and its share is next to the Microsoft leading competitor. This company has variety of services and largest data center that continues to provide facilities to billions of customers. AWS is very well-known public cloud that offers many services to connect for installations to the cloud. It also serves everything like disaster recovery and burst capacity [17].

       Google Cloud Anthos: The Google Cloud Platform is another one popular in hybrid cloud. It is a competitor of Microsoft AWS and IBM. Google primarily made pure cloud system, but later, they changed policy and started to work with on-premise systems for disaster recovery, elastic infrastructure, Big Data, and DevOps. It also provides a huge number of cloud-based services. The services are based on AI efforts based on AI processor and TensorFlow. No one can buy TensorFlow system but can run AI and machine learning apps on Google Cloud [17, 18].

       Oracle Cloud at Customer: This is another one popular hybrid cloud service provider. It provides mostly on-demand service, in its own cloud system. Unlike Azure, AWS, and GCP, this provider does not allow its software to execute in virtual instances for any operation. But it runs on metal servers; Oracle also offers this kind of service. Oracle cloud is also very easy to run its apps on-premise on the cloud [18].

       IBM: IBM merged all of its cloud services, called IBM Cloud. It possesses more than 170 types of services for public cloud and on-premise. These services are not only limited to bare metal hosting and virtualized mode, containers, and server less computing, DevOps, AI/ML, HPC, and blockchain. It also offers to do lift and shift on-premise apps, executing on IBM platforms [18].

       Cisco Cloud Center: Cisco is popular for private cloud that also offers hybrid solutions via its partner. Cisco Cloud Center is more secured to manage and deploy the applications in different data centers in both private and public cloud environments. Cisco’s partner networks are Google, CDW, Accenture, and AT&T. Google is the biggest partner among them. It offers the hybrid connectivity and their solutions [18].

       VMware vCloud Suite: VMware provides vendor for virtualized services. It is relatively new than that of other service providers. VMware has the vSphere hypervisor. Customers can run in some known public clouds or their own data centers or cloud provider partners. These cloud providers are able to run vSphere on-premise that creates a stable hybrid cloud infrastructure [19].

      1.3.1 Cloud Storage and Backup Benefits

      Protection of the confidential data is very difficult. Automatic backup of cloud storage is flexible. It also provides data security.

      Microsoft Azure is very effective in SaaS. Whereas, Google Cloud is strong in AI [18]. Table 1.1 gives a comparison among them.

      1.3.2 Pros and Cons of Different Service Providers

      All the cloud service providers have their own pros and cons. Their make themselves a suitable choice for different purposes. Here, the advantages and disadvantages are described for all the providers. Table 1.2 provides a comparative study on this.

      Table 1.1 Comparison between AWS Outpost, Microsoft Azure Stack, and Google Cloud Anthos.

AWS Outpost	Microsoft Azure Stack	Google Cloud Anthos
Amazon has a huge tool set and that too is rapidly growing. No service providers can match with it. But the pricing is bit puzzling. Though providing service for hybrid or public cloud is not amazon’s primary focus thus incorporation of cloud services with on-premise data is not in top priority [20]. They primarily focus on public cloud.	The customer can run in their own data center. Azure tries to incorporate with that. It provides the facility of hybrid cloud [19]. A customer can replicate his environment in Azure Stack. This is very useful in case of backup disaster and for cutting cost.	Google has come to the cloud market later. So, it does not have that much level of focus to incorporate the customers. But the strength is its technical efficiency. Some of its efficient tools are applicable in data analytics, machine learning, and deep learning.

      Table 1.2 Pros and cons between AWS Outpost, Microsoft Azure Stack, and Google Cloud Anthos.

Vendor	Strength	Weakness
AWS Outpost	Dominant market positionExtensive, mature offeringsEffective use in large organizations	Managing costVery difficult for usingOptions are overwhelming
Microsoft Azure Stack	Second largest service providerCoupling with Microsoft softwareSet of features is vastProvides Скачать книгу В начало < 9 10 11 12 13 14 15 16 17 18 > В конец