Application Scenario: Website Application
Scenario
Individuals or businesses building portal websites, office OA, and other moderately loaded sites need to ensure smooth business operations and a stable, reliable system.
Pain Points
Traditional servers are costly, with significant expenses and a tendency to lead to resource wastage;
Traffic spikes unpredictably, and without sufficient elastic scaling, existing servers may become overloaded, leading to poor user experience.
Recommended Applications
During the initial stage of a website, when traffic is low, a single low-specification elastic cloud server instance can run web applications like Apache or Nginx, databases, and store files;
As your website grows, you can upgrade the configuration of your elastic cloud server instance or increase the number of instances at any time, without the concern of insufficient resources during business surges due to low-specification computing units.
Product Advantages
eSurfing Cloud's elastic cloud servers are ready to use immediately, with delivery in minutes, allowing for rapid access and deployment in data centers across the country;
You can change specifications and scale up as needed without disrupting business operations, balancing cost and ensuring a good user experience;
By using load balancing, distribute the workload across multiple cloud servers to enhance business processing capabilities;
By leveraging the high concurrency support of Object-Oriented Storage, resolve the issue of webpage crashes during frequent website visits.
Recommended combinations
Autoscaling service, elastic load balancing, cloud database, object-Oriented Storage.
Scenario architecture diagram
Application Scenario: E-commerce
Scenario
In scenarios such as promotions, flash sales, bestsellers, and livestreaming, the traffic of e-commerce websites may fluctuate dramatically in a short period of time. This requires servers to have high performance, rapid elasticity, and high stability.
Pain Points
Livestreaming and flash sales activities have high demands on website performance; slow system response can affect the user experience;
When traffic is high, slow page loading and large network latency can occur, while at times of low traffic, there is a risk of resource wastage.
Recommended Applications
Autoscaling can be used in conjunction to automatically increase the number of elastic cloud server instances before the peak in requests arrives, and decrease the instances during periods of low demand. Meet the resource requirements when traffic reaches its peak while reducing costs.
Product Advantages
Automatically adjust elastic computing resources to handle business peaks at any time;
Utilize object-Oriented Storage technology to store streaming data or file data in the cloud, addressing the pressure of high throughput and high concurrency business needs;
For situations with high traffic, load balancing can be used to distribute the workload across multiple cloud servers to ensure a good user experience.
Recommended combinations
Autoscaling service, elastic load balancing, object-Oriented Storage.
Scenario architecture diagram
Application Scenario: Game Deployment
Scenario
Game services have extremely high requirements for server performance, reliability, networking, and autoscaling capabilities.
Pain Points
Insufficient server performance can lead to serious issues such as lag and service interruptions, which can affect the player's gaming experience and cause user churn;
Inability to accurately predict development cycles, player numbers, and growth rates can easily lead to insufficient resources or resource wastage.
Recommended Applications
Elastic cloud servers provide high-reliability, high-performance, high-stability, and low-latency computing support for gaming services, offering players a smooth gaming experience;
Rapid elastic expansion according to the needs of the gaming business, making rational use of computing resources to save costs.
Product Advantages
Multiple nodes and Availability Zones help game services quickly achieve a wide range of deployment;
Elastically scale resources quickly according to demand, use resources wisely, and save costs;
Professional GPU virtualization technology and process-level resource scheduling provide greater concurrency per server, reducing the cost of cloud adoption;
Advanced encoding and decoding protocols, combined with optimized network transmission, ensure lower latency and better experience.
Recommended combinations
Auto Scaling, elastic load balancing, cloud database.
Scenario architecture diagram
Application Scenario: Big Data Analysis
Scenario
Big data analysis scenarios involve processing large volumes of data, requiring high I/O capabilities and rapid data exchange processing capabilities. For example, compute-intensive tasks like MapReduce and Hadoop.
Pain Points
The volume of business data is continuously increasing, which demands substantial computing resources, resulting in cost pressure for enterprises;
Data processing is periodic, with frequent business fluctuations, requiring flexible resource scheduling.
Recommended Applications
Using cloud servers with local disks ensures a vast storage space and high storage performance, while also providing higher network performance for Hadoop and Spark clusters in the cloud.
Product Advantages
Using cloud servers with local disks provides highly reliable storage resources, facilitating the secure storage of data, files, and applications;
For key protected data, the multi-redundancy feature of object-Oriented Storage can be used to achieve disaster recovery for data in different locations;
You can expand online elastically on demand without the need to migrate data, meeting the challenge of large capacity and high concurrency requests.
Recommended combinations
Autoscaling service, elastic load balancing, object-Oriented Storage.
Scenario architecture diagram
Application Scenario: Graphic Rendering
Scenario
Professional-level CAD, video rendering, graphic processing, and other scenarios require powerful computing capabilities, pursuit of ultimate performance experience, and a super cost-performance ratio.
Pain Points
Long rendering times: A single special effects shot can take more than ten hours to render;
Low rendering efficiency: With limited equipment, a large number of rendering tasks have to be queued for execution;
High rendering costs: Upgrading and maintaining physical machines consume a significant amount of operational costs.
Recommended Applications
Heterogeneous GPUs have excellent GPU computing acceleration capabilities, allowing for real-time rendering of images within seconds, widely used in scenarios such as graphic rendering, cloud-based graphic workstations, and video transcoding.
Product Advantages
Utilizing cutting-edge GPU hardware from the industry and keeping in sync with the latest technologies, it allows for seamless switching to the newest hardware models;
Powerful functionality, fully supporting a variety of GPU applications, deep learning frameworks, such as OpenGL, DirectX;
Convenient and fast, providing the same usage methods and management features as standard cloud servers;
Rich specifications, offering a variety of video memory options to meet different graphics and image processing scenarios;
Monitor and alert GPU resources through cloud eye services.
Recommended combinations
Autoscaling service, cloud eye.
Scenario architecture diagram
Application Scenario: Deep Learning
Scenario
Deep learning training or prediction platforms require a significant amount of computing resources and pursue excellent performance.
Pain Points
The deep learning process generates a large amount of temporary data, which has high requirements for storage;
Deep learning neural network computation has extremely high requirements for network latency.
Recommended Applications
For deep learning scenarios involving continuous and extensive artificial neural network computations, eSurfing Cloud recommends GPU instances, which not only offer outstanding performance but also significantly save on costs.
Product Advantages
By using object-Oriented Storage, cloud databases, cloud eye, and other services, you can quickly build a fully functional deep learning offline training system;
Powerful functionality, fully supporting a variety of GPU applications, deep learning frameworks, such as CUDA, OpenCL;
Convenient and fast, providing the same usage methods and management features as standard cloud servers;
Rich specifications, offering a variety of video memory options to meet different graphics and image processing scenarios;
Monitor and alert GPU resources through cloud eye services.
Recommended combinations
GPU Cloud Servers, object-Oriented Storage, Cloud Database, Cloud Eye.
Scenario architecture diagram
Application Scenario: Self-built Database
Scenario
Database scenarios have high memory requirements and need to meet the demands of massive data.
Pain Points
Large amounts of data require high memory specifications for cloud servers, leading to expensive equipment procurement costs;
High volume of data access requires fast data exchange and processing, and traditional methods cannot scale up or down in a timely manner.
Recommended Applications
On-demand usage allows you to purchase memory-optimized cloud servers of different specifications based on business needs, saving costs;
Configure a Elastic Volume Service with ultra-high IO and appropriate bandwidth, with elastic expansion capabilities to quickly process massive amounts of data.
Product Advantages
Ultra-high I/O EVS have strong read and write performance and high throughput bandwidth, which can meet the deployment needs of various user databases (such as MySQL, NoSQL);
It can notify the results of autoscaling strategies in real time, allowing users to understand the dynamics of scaling and adjust strategies in a timely manner to ensure the stability of database services;
Rapidly achieve network access and interconnectivity, with multiple replicas ensuring data security.
Recommended combinations
Autoscaling service, EVS.
Scenario architecture diagram
