There is no doubt that as a brand-new architecture, Serverless is considered to be the next generation of cloud computing innovative technology after virtual machines and containers, and it truly embodies the "essence" of cloud computing, that is, by better sharing the underlying Hardware resources improve utilization efficiency and allow customers to enjoy lower costs.
It can be seen that Serverless shields hardware physical resources, allowing services to quickly expand according to business needs, making it more flexible and elastic, and has built-in availability and fault tolerance, built-in service integration, and built-in security best practices; while enterprises tend to adopt Serverless for cloud migration. The key reasons are that during use, there is no need to configure or manage infrastructure; no need to install and maintain software components; and the ability to automatically expand based on consumption units, etc.
As the pioneer and leader in the serverless field, Amazon Cloud Technology officially released Amazon Lambda as early as 2014, which can be said to have launched the first wave of serverless commercialization. In the past nine years, Amazon has Cloud Technology continues to maintain the "pace" of innovation in the serverless field, and makes significant progress in the serverless field almost every year. As the saying goes, "accumulating silicon can lead to thousands of miles." Today, its serverless services have covered computing, storage, network, containers, Databases and multiple aspects of integration truly make Serverless “normal”.
At the 2023 re:Invent Global Conference, Amazon Cloud Technology not only released three Serverless innovations in the database and application fields, but also elaborated on the construction of Amazon Cloud Technology Serverless for the first time with the theme of "Road to Serverless". The road ahead, as well as the exploration and efforts behind it, and this attitude and spirit of continuous innovation in Serverless are the foundation and confidence behind Amazon Cloud Technology’s strong “dominance” in the Serverless field today.
Adhere to innovation and accumulate silicon to reach thousands of miles
If before 2017, Amazon Cloud Technology’s innovations in the Serverless field were mainly concentrated in infrastructure fields such as computing and storage, then starting from 2017, Amazon Cloud Technology’s technological innovations in the Serverless field were mainly concentrated in data and data analysis services. field, among which Amazon Aurora's "Serverless" has "epoch-making" significance in terms of the innovation of Serverless technology-empowered databases.
The reason is that Amazon Aurora Serverless, released in 2018, is the first relational database service in the industry to release Serverless features. Compared with open source hosted databases, its cost performance is 3-5 times higher, so it not only changes traditional database products At the same time, it provides more efficient, more flexible, safer and more reliable data management solutions for enterprises' "cloud native data infrastructure".
For this reason, at the recently held 2023 Amazon Cloud Technology re:Invent China Tour event, Chen Xiaojian, General Manager of Amazon Cloud Technology Greater China Product Department< a i=2> I am proud to say that the launch of Amazon Aurora Serverles and its huge influence are actually inseparable from several key "secret weapons".
The first major weapon is the support provided by Grover.Grover has "created" two new concepts in the field of distributed databases. One is that logs are data, and the use of traditional Compared with the method of synchronizing data across nodes, Grover will only synchronize logs between nodes, which has the advantage of greatly reducing the bandwidth required by the network and the cost of I/O reading and writing of the database disk; secondly, it is the first to synchronize logs between nodes. The data and computing resources of the database are separated, enabling serverless expansion of database storage, which lays a solid foundation for large-scale implementation of Serverles.
The second largest weapon is Caspian (collaborative resource management system), which implements the allocation of "dynamic resources" in the database. Caspian is an innovative virtual machine monitor, thermal management system, and a system that updates and combines the database engine itself. The "blessing" of these technologies allows the Amazon Aurora database to adapt to changes in load. It can quickly adjust the memory capacity in milliseconds, thereby quickly responding to changes in database workload. In other words, it has the ability to "move resources" in real time, and the entire process will not have any impact on the actual performance of the database instance.
The third major weapon is database sharding. The core problem it solves is that the instance application of the database exceeds the physical limitations of the server. Currently, sharding has become a common technology to improve database performance. Through horizontal splitting, it can split an instance into multiple physically distributed database servers. , but when users use it, they still use it according to a logical instance of the database. But this process is very complicated, including how to manage custom routing and coordination layers, how to rationally re-shard and manage scaling, how to achieve "consistency" in different isolations, etc. To this end, Amazon Cloud Technology has also launched a truly "sharded" Amazon Aurora service through continuous technological innovation.
The fourth major weapon is the time synchronization infrastructure (Amazon Time Sync Service).For the database, time synchronization is crucial, because the database is important for data Read and write guarantees, especially in relational databases that support things, need to be achieved by strictly following the date order of operations. This is very easy to implement on a machine, but for distributed databases, how to ensure that " It is "extremely difficult" to achieve time synchronization among hundreds or thousands of server nodes.
To this end, Amazon Cloud Technology uses accurate clocks and compares "timestamps" to determine the order of operations of many servers, thereby ensuring time consistency. Therefore, the accuracy of the clock becomes very important, and this is also the time This is the context in which the synchronization infrastructure "came into being", based on the latest generation of Amazon Nitro chips that support custom hardware that accurately synchronizes their local clocks according to time, with pulses transmitted by a custom time synchronization network. Meanwhile, in such a cabinet, on top of its rack is a specialized reference clock, with precise time derived from satellite-based atomic clocks, thereby providing synchronized clocks accurate to a billionth of a second error anywhere in the world. In addition, each rack is equipped with a local atomic clock, in case the satellite to which it is synchronized is temporarily unavailable.
More importantly, every step of this is done through hardware, without the need for drivers, operating systems or network buffers. In other words, the time synchronization infrastructure is built on the dedicated Amazon Cloud Technology network infrastructure. A high-precision time synchronization service that only does "time synchronization".
Continue to lead, the foundation and confidence behind it
In fact, from the four "secret weapons" behind Amazon Aurora Serverless, it is not difficult to see that Amazon Cloud Technology can continue to iterate and evolve in the serverless field. The technological innovation spirit of "Never Stop" has played an important value and role.
Likewise, this also lays a key foundation for Amazon Cloud Technology to continue to lead the path of serverless innovation. At the 2023 re:lnvent global conference, Amazon Cloud Technology once again launched three Serverless services to help customers analyze and manage data at any scale and significantly simplify operations. Customers do not need to spend time and energy to configure, manage and expand their data. Infrastructure, specifically:
First, there's the Amazon Aurora Limitless database, which automatically distributes and queries data across multiple Amazon Aurora Serverless instances and can scale to millions of transaction-level writes per second and manage petabytes of data, and here's the key The aforementioned “sharding” technology.
It is understood that the Amazon Aurora Limitless database provides a single interface and automatically expands and replicates data across multiple database shards. Not only that, Amazon Cloud Technology also uses Amazon Aurora Database itself to build a fast routing layer to efficiently route queries to the appropriate shards and achieve fast distributed data exchange among all shards.
Chen Xiaojian emphasized that the launch of Amazon Aurora Limitless database has important value in two aspects: on the one hand, it is the only sharded database in the industry that truly uses serverless implementation. Users will no longer have to manually manage the underlying physical servers. Resources, and just through this service, you can automatically realize resource management and the ability to expand and shrink; on the other hand, this also means that users will be able to truly achieve no expansion "upper limit" in the future. Distributed database.
Second, there is Amazon ElastiCache Serverless, a service that helps users create highly available caches in a minute and scale vertically and horizontally in real time to support customers' complex applications without the need to manage infrastructure.
Users can use Amazon ElastiCache Serverless to run caching for demanding workloads without spending time on capacity planning or requiring caching expertise. At the same time, Amazon ElastiCache Serverless continuously monitors an application's memory, CPU, and network resource utilization and instantly scales to adapt to changes in the access patterns of the workloads it serves. Not only that, users can also create a highly available cache where data is automatically replicated across multiple Availability Zones and provides an availability service level agreement (SLA) of up to 99.99% for all workloads, saving users time and money. . Likewise, Amazon ElastiCache Serverless is compatible with two popular open source caching solutions, Redis and Memcached, without the need to perform resource infrastructure management tasks.
Finally, Amazon Redshift Serverless has added a new feature, which can use artificial intelligence (AI) to predict workloads and automatically expand and optimize resources to help customers achieve cost-effective goals.
Includes a predictive model based on machine learning that can predict future workload patterns and adjust resource capacity in advance; with a real-time query analyzer, and using machine learning to estimate the resource requirements of each query and make appropriate allocations, the system can analyze each query has more than 50 unique features; in addition, according to user needs, each query can be optimized to reduce costs or improve performance, and can query different expansion modes such as linear, sublinear and superlinear. Through the powerful "blessing" of AI, it can Let Amazon Redshift Serverless be optimized to handle different types of complex production workloads, large and small, improving price/performance by up to 10x.
It can be seen that with the addition of the three newly released Serverless services, it can be said that Amazon Cloud Technology has added "strong players" to its product lineup in the field of full-stack Serverless data services, and this is why Amazon Cloud Technology continues to lead Serverless innovation It is an important manifestation of the foundation and confidence of evolution.
Reshape again and make Serverless “normal”
It can be seen that through years of innovation experience and technology accumulation, Amazon Cloud Technology has played an important role in helping enterprises and developers around the world use Serverless services to reduce operation and development costs and increase agility and resilience. It can be said that in the field of Serverless, Amazon Cloud Technology has not only achieved comprehensive leadership, but also has vigorously promoted Serverless from being "unattainable" before to "flying into the ordinary" by continuously deepening Serverless's comprehensive layout and underlying innovation in cloud services. "People's Home" has truly realized the "normalization" of Serverless. We can make further observations from several dimensions:
First, Amazon Cloud Technology has both “depth and breadth” in the Serverless field and can provide rich full-stack Serverless cloud services and end-to-end data capabilities. Among them, from a breadth perspective, Amazon Cloud Technology’s Serverless services widely cover computing, storage, networks, containers, databases, data analysis and other aspects; from a depth perspective, Amazon Cloud Technology continues to Refine the functions of Serverless services and continuously improve their performance. This full-stack "Serverless" capability not only confirms Amazon Cloud Technology's "dominance" in Serverless from one side, but also enables Serverless to achieve "normalcy" The key foundation of ”.
Second, Amazon Cloud Technology continues to maintain market leadership in the Serverless field, not only because it always adheres to six important cloud computing service attributes in the Serverless field: security, availability, scalability, performance efficiency, and cost. efficiency and sustainability. What’s more important is that Amazon Cloud Technology has truly penetrated into the “lowest” core technology innovation in the Serverless field, and this capability is “far ahead” in the world.
For example, in addition to the four "secret weapons" behind Amazon Aurora Serverless mentioned above, it is also reflected in the technical foundation core "Firecracker" in Amazon Lambda, which is a secure and fast micro-virtual machine for serverless computing. , with better resource utilization, faster startup time, etc.; another example is "Amazon Lambda SnapStart", which uses "Firecracker" to improve the cold start of Java-based functions, and users can start the function without changing the code. 10 times faster; in addition, "Serverless on Graviton" also makes Amazon Cloud Technology more cost-effective and so on. From this perspective, it is Amazon Cloud Technology’s continuous innovation at the “lowest level” in the field of Serverless technology that has enabled Serverless to achieve milestone evolutions again and again.
Third, technological innovation in the serverless field also allows Amazon Cloud Technology to operate its global infrastructure more effectively and reduce idle resources, which also means it is more "green and low-carbon." Similarly, these technological innovations can also empower more industry customers and maximize the technological dividends brought by Serverless innovation.
In other words, in the future, with the help of Amazon Cloud Technology's rich full-stack Serverless service portfolio, users from hundreds of industries will be able to carry out various innovations based on Amazon Cloud Technology's Serverless services, which will not only further lower the application threshold of Serverless, but also obtain practical results. Real benefits will accelerate the pace of in-depth "cloud use" and digital intelligence transformation.
Generally speaking, the emergence of cloud computing breaks the boundaries of the underlying infrastructure and makes innovation easier; while the emergence of Serverless breaks the boundaries of the cloud and makes innovation more efficient. In this process, Amazon Cloud Technology has once again liberated the virtual infrastructure and cloud data and analysis service areas on the cloud through the continuous innovation of Serverless and the attitude and spirit of "taking small steps to reach a thousand miles", releasing It not only defines and builds the innovation path of Serverless, but also has the ability and confidence to lead the continuous innovation of Serverless and promote the "normalization" of Serverless.
Click "Read the original text" to quickly view all the hot releases of Amazon Cloud Technology re:Invent 2023!
Shenyao's Technology Observation was founded by Shensky, a senior technology media person. He has 20 years of experience in enterprise-level technology content communication and has long been focused on the observation and thinking of industrial Internet, enterprise digitalization, ICT infrastructure, automotive technology and other contents.
