With the digital transformation of thousands of industries, the amount of data has skyrocketed, and data has gained unprecedented attention. As one of the three major IT infrastructures for storage computing networks, data storage has always accounted for a significant portion of IT spending. However, under the trend and pressure of the epidemic and economic downturn, how to reduce costs and increase efficiency through innovative IT technology has become a problem that enterprises and technical teams must face.
When it comes to flexible cloud resources and storage, the industry will inevitably think of Amazon S3, the ace storage product of Amazon Cloud Technology. In the 16 years since the launch of S3 in 2006, Amazon Cloud Technology has been investing in research and development in storage technology. For example, in 2021 alone, more than 130 new services and features in the storage field have been released. The optimization of storage cost and performance by Amazon Cloud Technology provides an important reference model for the industry.
For 16 consecutive years, the cost of storage has continued to decline. How to do it?
In terms of storage technology, in order to improve data storage and recall performance and reduce the cost of storage media, it is usually an important part of storage technology to use the characteristics of various storage media to classify storage according to the importance of data. The price of Amazon S3, the ace storage product of Amazon Cloud Technology, has been reduced by 7 times in 16 years (take the S3 standard frequent access tier Amazon S3 Standard Frequent Access as an example). By the end of 2021, Amazon S3's standard infrequent access tier and single-zone infrequent access tier Another 31% off the price of storage!
The confidence to compress the storage cost space is the design logic behind the tiered storage technology of Amazon Cloud Technology.
The amount of data has exploded, and the complex data can actually be subdivided into various types, such as real-time trading platform data, business monitoring data, data involved in machine learning inference, and social media generated data, audio, video and image data. In addition, disaster recovery is required. Backups, development test data, historical archive data, and more. Amazon Cloud Technology defines the "data temperature" for these data, which is divided into hot data, warm data, cold data and frozen data according to the frequency of data calls.
Figure: Massive data and multiple data access requirements
In 2018, Amazon Cloud Technology launched the Amazon S3 intelligent tiering function for massive data with unknown access patterns. Amazon S3 can automatically tier hot and cold data for up to 68% cost savings.
In addition, Amazon cloud technology data tiered storage has been extended to file storage Amazon EFS. Amazon EFS offers four storage tiers that save up to 72% on storage costs.
Now, Amazon S3 intelligent tiering has covered almost all storage categories of the Amazon S3 family, enabling intelligent storage and use of data throughout its life cycle. Users can use Amazon S3 Smart Tiering as the default storage tier to build various cloud applications including data lakes and machine learning.
Storage moves to scene segmentation, Amazon cloud technology accumulates storage products across the board
When it comes to storage application scenarios, digitalization has penetrated into various industries and has also spawned a variety of business scenarios. According to the development process of cloud computing, we can divide business scenarios into cloud-native scenarios and traditional enterprise business scenarios.
Such as website hosting, data lake, digital marketing, cloud photo album, audio and video sharing, e-commerce platform, social media and other applications, these cloud-native modern applications, customers can use cloud-native serverless file storage Amazon EFS, get simple, no management cost, Massively elastic expansion, highly integrated, automatically optimized experience.
In addition, for more complex and diverse enterprise-level applications such as ERP/CRM, database, autonomous driving, life science, hybrid cloud, simulation, meteorological analysis, EDA, video rendering, etc., Amazon Cloud Technology proposes the idea of "dedicated construction" to Provide cloud storage products and services.
For enterprise-class workloads, Amazon Cloud Technologies offers the Amazon EBS IO2 BX and Amazon FSx series of services. Amazon EBS IO2 BX is designed for applications running on SAN storage in offline data centers. The Amazon FSx family is purpose-built shared file storage for a variety of enterprise applications. FsX = File X System, in detail, the Amazon FSx family already has the following storage products:
-
Amazon FSx for Windows File Server , launched in November 2018, supports Windows Server workloads.
-
The new generation of Amazon FSx for Lustre , released in November 2021, is aimed at HPC computing-intensive workloads , helping customers quickly complete high-performance computing, such as gene sequencing, new drug development, autonomous driving simulation, semiconductor design and other tasks.
(1) Amazon FSx for Lustre provides 5x the throughput (up to 1GB/sec per TB) compared to the current generation.
(2) Amazon FSx for Lustre supports two-way synchronization of data updates with S3. When data is updated on either side of S3 or Lustre, automatic and fast synchronization can be achieved on the other side.
-
Amazon FSx for OpenZFS , November 2021 release for big data analytics Linux workloads . For cloud migration customers who previously put their data on ZFS or other Linux file servers and don't want to change the architecture. It allows ZFS or other Linux-based file servers to be migrated to Amazon Cloud Technology without changing the way they apply and manage data, delivering up to 1 million IOPS with a latency of just a few hundred microseconds, making it ideal for machines running like Learning, audio and video process these data-intensive workloads in real time.
-
Amazon FSx for NetApp ONTAP , released in September 2021, facilitates the seamless migration of enterprise businesses to the cloud. Enables customers to start, run, and scale fully managed NetApp ONTAP file storage services on Amazon Cloud without modifying their applications, tools, processes, or workflows to migrate their applications running on NAS devices to Amazon Cloud . Amazon FSx for NetApp ONTAP enables customers to deploy and manage applications and data using the NAS workflows they are already familiar with, while enjoying the agility, scalability, and security of Amazon Cloud, combined with other Amazon Cloud services seamless integration between.
Through the subdivision of scenarios and research on corresponding storage requirements, Amazon Cloud Technology has covered a full line of storage products and services from block storage, object storage, to file storage, backup, disaster recovery, as well as data transmission and edge processing. The "family bucket" product matrix of cloud giant Amazon Cloud Technology in the field of storage technology.
A powerful product matrix, covering a comprehensive scene, is not a simple propaganda slogan. For example, the scene covered by Amazon's cloud technology storage product "Family Bucket" has even been extended to the International Space Station outside the earth. This year, Amazon Cloud Technology has deployed the Amazon Snowcone service to the International Space Station. The device contains a solid-state drive, which reduces the end-to-end data download and analysis events for the International Space Station from 20 hours to 20 minutes.
This should be regarded as an extreme edge scenario. After all, compared with the scenario on Earth, in the edge scenario of the International Space Station, the hardware must be as light as possible, withstand extreme vibration and gravity during launch, and safely connect to the host in microgravity. Spacecraft power, cooling and networking systems.
Figure: Amazon Snowcone "serving" on the International Space Station
From a storage technology perspective, the Amazon Snowcone is designed to run edge computing workloads, all protected by multiple layers of encryption. After the data is collected and processed locally, the device is usually shipped back to an Amazon cloud technology data center so that the processed data can be stored in the cloud and further processed. Alternatively, Amazon DataSync can be used to copy data from the Snowcone device back to the Amazon Cloud Technologies data center.
In addition to extreme edge scenarios, take the case of Jinfeng Huineng in the energy and power scenario as an example.
Summarize
Behind every generation of scientific and technological iterations and changes, it is because technology has effectively improved the efficiency of people's production and life. This is also the ultimate value of IT swallowing everything and reshaping thousands of industries. In the uncertain and changing market environment, how to continuously reduce costs and improve efficiency is a problem that every technical person should think about. In technical practice and accumulation, Amazon Cloud Technology's product construction ideas in the storage field have provided a good reference sample for the industry, which is worthy of reference and reference.
Of course, Amazon Cloud Technology has a rich matrix of storage products and services, as well as best practices, application cases, and many more. From September 22 to 23, 2022, Amazon Cloud Technology will hold an important technology event in Shanghai - the 2022 Amazon Cloud Technology China Summit. Click the registration link or scan the QR code in the picture to get more information about the conference , and more technological innovations are waiting for you to discover.
Registration link:
