Preface
As of this year (2022), Amazon Cloud has been in its 16th year and has been recognized by Gartner as a cloud computing leader for eleven consecutive years. In terms of EC2, it has grown from the original single instance to today supporting 475+ instance types, and in terms of service types, it has developed and extended to all walks of life. Speaking of my relationship with Amazon Cloud Technology, I have to start from when I was in college, from experiencing the service resources on Amazon Cloud for free back then to my continuous learning on Amazon Cloud today. Next, I will talk about the learning process on Amazon Cloud from the aspects of product use, participation in innovation conferences, and in-depth reading of technical resources.
Start my cloud service with a free trial of Amazon Cloud
For a cloud developer, the free plan is a considerable attraction of AWS compared to many other cloud vendors. Its free package covers more than 100 products in 20 categories, including computing, analysis, database, Internet of Things, machine learning, storage, etc.[1] Of course, different categories correspond to different industry groups, and I am more concerned about computing, database and storage services. Taking computing and storage as an example, the computing resource Amazon EC2 provides 750 hours/month of free computing services for 12 consecutive months in the first year, and the storage resource Amazon S3 provides 5GB of free standard storage capacity for the first 12 months. The free package gave me a better understanding of Amazon Cloud when I was in college. I was amazed at how fast it was to create instances and how quickly the components were available. Naturally, I was still using Amazon Cloud products after the free period ended.
As a big data practitioner, I also like to "look for trouble". Since I maintain a personal blog website, I need to record and count visitors from all over the world in real time to understand the popularity of the website and articles. Then the data lake storage service built on Amazon S3 is my choice. From a technical perspective, I chose Amazon cloud products as the base of my data lake because Amazon S3 provides 11 nines of durability (99.999999999%) and four nines of availability (99.99%), and is used as an object storage The leader, the breadth of its interface has been supported by a number of other object storage providers such as minio, which means that my underlying storage can be developed in one place and can be used in multiple places. At the same time, create the Flink engine and Presto operating environment on the AWS EC2 instance, and you can write the data in real time to the hudi data lake built on Amazon S3 and use Presto for hotspot query display.
Actively participate in the Amazon Cloud Technology Innovation Conference to learn about future technology development directions
As a pioneer of cloud service providers, Amazon every technological innovation brings innovation and upgrade to our business. The Amazon Cloud Technology Innovation Conference is an online conference regularly held by Amazon Cloud Technology. It aims to comprehensively demonstrate AWS's technological innovation capabilities and customer best practices, and bring users a new perspective on using AWS. It is a topic that developers must pay attention to. In March 2021, at the Amazon Cloud Technology AI Online Conference, Amazon Cloud Technology joined hands with nearly 30 customers and partners to reveal to you how artificial intelligence and machine learning (AI/ML) empower innovation, drive transformation, and the impact of AI in various industries. /ML field observations. In October of the same year, again with the theme of "data-driven innovation", the modern data strategy trilogy was disclosed for the first time, in-depth analysis of the intelligent lake warehouse architecture to achieve free flow of data, a panoramic interpretation of Amazon Cloud Technology's full-stack data analysis service, and sharing of typical data-driven The way to innovate enterprises. Each meeting brings different technical insights. [3][4][5][6]
Amazon Cloud Technology also launched an innovation conference with the theme of cloud infrastructure on April 20 this year as scheduled, and I also actively participated in it. This conference is divided into one main venue and six sub-venues. The keynote speech in the main venue is the classic use case of Amazon Cloud Technology customers, relying on Amazon Cloud to bring cloud experience to enterprises such as Daewoo Infinity and Transsion Holdings. The six sub-venues were themed on reconfiguration of computing, cloud load, data innovation, borderless cloud, cloud security and best transformation practices. As in previous sessions, there will be video replays of the keynote speeches at each branch venue, so there is no need to worry about missing the keynote speeches that will be broadcast live at the same time. At the same time, this conference also provided resources such as product cards, training and certification, and data centers, allowing us to better understand the power of Amazon Cloud technology [7].
At this year's Innovation Conference, what impressed me most was the pace of EC2 innovation mentioned in the keynote speech, which just solved my confusion about how fast Amazon Cloud EC2 instances can be created and how the performance is almost indistinguishable from bare metal. As General Manager Gu Fan said, Amazon Cloud EC2 has made the following four innovations on the research and development road, namely Nitro architecture, self-developed chip design, accelerated computing, and container/serverless.
In the early days, AWS also used software virtualization to divide physical server resources into multiple virtual instances to provide EC2 instances, but this solution also had certain problems. Software virtualization will consume 30% of the performance of the hardware. At the same time, the virtual machine can also access the underlying hardware resources, which poses certain security risks. Moreover, if the iteration of storage and network virtualization cannot be decoupled from the server hardware itself, it will be difficult to accelerate EC2 instances. research and development process. The Amazon Nitro architecture completely solves the above problems. It offloads virtualization to the hardware level to minimize performance loss; in terms of security, it completely isolates the underlying write operations of user instances, thereby improving security; and A variety of Nitro network cards and memory cards can virtualize network and storage and completely decouple server hardware updates, which can greatly speed up the development process of EC2 instances.
Chip innovation is an important pillar in reconstructing the cloud base. AWS has been developing self-developed chips for the cloud since 2013. It works backwards based on users' deep understanding of workloads in the cloud to design more suitable cloud chips. on the chip. Over the years, AWS has developed three product lines, including the Nitro series, the machine learning and reasoning series, and the Graviton series based on ARM architecture. As the leader among them, the Graviton series brings the best price/performance ratio, enhanced security and a broader ecosystem.
AWS has the most extensive range of machine learning instances, and its self-developed INF1 chip has greatly improved throughput and cost compared to similar instances from NVIDIA. Take OPPO's use of INF1 to enhance conversational AI-Xiaobu Assistant as an example, which can provide AI capabilities for semantic understanding and conversational interaction. The OPPO team compiled and optimized Xiaobu's NLP model for INF1 and deployed it on the INF1 instance for online inference. In two test scenarios of Q&A and chatting, INF1 saved 35% on the overall cost of NVIDIA T4. The end-to-end delay was reduced by 25%, and the throughput was increased by 2.3 times.
Relying on Amazon e-commerce, Amazon Cloud has the largest application base, which promotes its launch of container/serverless computing architecture. With the increase in business scale, Amazon e-commerce has evolved from the initial single application to an SOA architecture, which is divided into commodity services, order services, user services and other systems based on functions. But as time goes by, each sub-service system becomes a smaller single application system. Business drives technological change. In 2006, Amazon began to decompose applications into multiple microservices, which truly made Amazon's innovation more flexible. Today, Amazon e-commerce is running more than 100,000 microservices.
Summarizing the development experience of Amazon’s e-commerce microservices, it is natural for Amazon Cloud Technology to launch containerized services. AWS's 16 years of continuous technological innovation in the serverless field have brought cloud user use to its peak. Provide users with an extraordinary user experience.
On April 15, Gartner released the 2021 global cloud computing market share report, and Amazon Cloud Technology ranked first with a share of 38.92%. Of course, this is inseparable from its deep involvement in cloud services and its insistence on continuous innovation. I also gained a lot from this year’s innovation conference. I gained a deeper understanding of the development history of AWS from the underlying technology, which allowed me to be more comfortable in future technology selections; I also changed my personal learning direction from the development of technology. , the application development trend based on containerization/serverless is the direction I should refer to and choose. After all, there are such high-quality cloud vendors, and it would be happier to focus entirely on business development!
Read more about Amazon cloud technology resources to improve your competitiveness
Of course, for me, in addition to using Amazon Cloud's rich products and watching cutting-edge technology innovation conferences, Amazon's knowledge architecture is also a guiding plan for my daily technology improvement. In the field of cutting-edge technology, its China Summit, Cloud Conference and Video Center are wonderful themes that we cannot miss. It has launched wonderful knowledge sharing videos for different engineer roles, different technical fields and different technical levels [4] [6] .
Through the above information, you can understand the current technology hot spots and development trends and prepare for the next step of learning. In terms of single technology, its complete product documentation allows us to make clearer choices about business technology selection and product development direction. Taking object storage as an example, through its detailed use cases and documentation, I learned how to build a data lake suitable for big data analysis, artificial intelligence, machine learning, and high-performance computing; how to determine the cost based on the hotness and coldness of the data. Tier storage to reduce operational complexity, gain new insights, and effectively reduce costs.
Summary and summary
For individual developers, Amazon Cloud is one of the best cloud service options because it almost covers the capabilities and services needed for personal development. In terms of learning platform construction, it has created a variety of learning platforms for developers:
- Getting Started Resource Center : Easily get started with cloud services from 0 to 1, covering: cost management, hands-on training, and development resources.
- Architecture Center : The Amazon Cloud Technology Architecture Center provides cloud platform reference architecture diagrams, vetted architecture solutions, Well-Architected best practices, patterns, icons, and more.
- Builder Library : Learn how Amazon Cloud Technologies builds and operates software.
- Build Tools : A toolkit for developing and managing applications on the Amazon Cloud Technology Platform
In terms of benefits, its free packages, discount gift packages and one-click deployment templates are not to be missed:
- Amazon Cloud Technology provides free packages for more than 100 products . Among them, the computing resource Amazon EC2 is free for 12 months in the first year, 750 hours/month; the storage resource Amazon S3 is free for 12 months in the first year, with a standard storage capacity of 5GB.
- Provide the latest discount gift package , 200 data and analysis discount coupons, 200 data and analysis discount coupons, 200Data and analysis coupons , 200 machine learning coupons, 200$ microservices and application development coupons .
- Solution CloudFormation deploys the template library with one click .
Whether it is entry-friendly welfare resources, detailed official documents, or regularly held technology innovation conferences, Amazon Cloud Technology can help you advance from a technical novice to a technical expert. This article introduces the growth of individual developers on the Amazon cloud technology platform from basic entry to advanced improvement. This is probably the reason why I insist on learning and using it! For many users, I think it will be a very good choice. For more details, please refer to the information in the reference resources.
Reference resources
[1] Amazon Cloud Free Plan registration address: https://aws.amazon.com/cn/free/?trk=c0722fb5-d267-4560-ba95-45a9d9890e54&sc_channel=el
[2] Introduction to Amazon Cloud S3: https://aws.amazon.com/cn/s3/?trk=044290d5-dfca-4a6b-8697-461c8f72ee8d&sc_channel=el
[3] Amazon Cloud Technology China Summit: https://www.awsevents.cn/summit/index.aspx?s=&smid=&vep_st=
[4] Amazon Cloud Technology Video Center: https://www.awsevents.cn/reinvent/video-center.html?s=&smid=&vep_st=
[5] Amazon Cloud Technology Cloud Conference: https://www.awsevents.cn/specialTopic/yuntus.html?vep_st=33533&s=&smid=
[6] Amazon Cloud online talk: https://aws.amazon.com/events/online-tech-talks/on-demand
[7] 2022 Amazon Cloud Technology Innovation Conference Cloud Infrastructure Summit: https://www.awsevents.cn/innovate/default.aspx?s=7286&smid=1269