Load balancing

load balancing

Kufan Cloud load balancing, a load balancing service that distributes traffic to multiple ECS servers, can expand the external service capabilities of the application system through traffic distribution, and improve the availability of the application system by eliminating single points of failure.

Product features

Exclusive instance performance A single instance has exclusive computing resources. The public IP address is independent and the number does not exceed
Support multiple protocols Supports Layer 4 (TCP/UDP) and Layer 7 (HTTP/HTTPS) protocol loads to meet the needs of different application scenarios
Improve ECS availability Forwards front-end access to two or more back-end business nodes, improving the availability of ECS
Easy to scale out Back end nodes can be increased or decreased freely according to actual scenario requirements
Applicable to services with high concurrent access High concurrency service access is evenly shared by multiple nodes, or bears different loads according to the specified weight value
Eliminate single node failure Multiple nodes are used to provide service access processing, thus avoiding the problem of service stagnation caused by single node failure

Product advantages

advantage
performance
usability
Scalability
cost
security

Kufan Cloud Load Balancing
The maximum concurrent number of a single cluster exceeds 10 million, which can handle the peak traffic of 20Gbps. You can safely deploy e-commerce websites, social app software, and game businesses with daily visits of more than 10 million to Kufan Cloud.
High availability, providing redundant backup, automatic failover, and ensuring that the back-end provides services normally, the cluster system will automatically eliminate the failed instances, and users only need to consider the back-end business code.
The cluster scales horizontally according to the business load to expand the external service capability of the application system, so that you don't need to worry about expansion. Users accessing resources in this process are transparent.
There is no need to purchase load balancing hardware, which greatly saves the cost of load balancing hardware by more than 90%.
Professional DDoS protection equipment is used to provide users with sophisticated anti DDOS attack capabilities for Internet applications, such as UDP, Flood, SYN Flood and CC attacks.

Traditional load balancing hardware
Traditional hardware devices, such as F5 BIG-IP and Netcaler, only support millions of connections, and the throughput is less than 5Gbps, which cannot meet the sudden and massive Web access requests such as "rush buying" and "second killing".
The dual host HA scheme provided by traditional hardware equipment is not enough to cope with the complex and changeable public network environment.
When traditional hardware devices, such as F5 BIG-IP and Netscape, encounter business bottlenecks, they need to re purchase hardware devices of higher specifications for redeployment, affecting business development.
The price of traditional hardware devices ranges from tens of thousands to hundreds of thousands per set, which is difficult to deploy; Professional operation and maintenance personnel are required, and hardware costs and labor costs have increased dramatically.
Traditional hardware equipment has limited ability to prevent DDoS intrusion

Open source solution
Self built schemes such as Nginx and HAProxy often have only tens of thousands of connections, which cannot cope with the massive concurrent requests of the Internet.
High requirements for technical capabilities, miscellaneous schemes, and availability can not be guaranteed.
Manual configuration requires manual intervention, with a long time cycle, great impact on business, and risk of operational errors.
The open source solution is free, but there is no unified standard, and the maintenance cost is high
Lack of mature security protection solutions.

Product Functions

Multiple protocol support

Rich protocol support to meet your diversified application scenarios

Specific protocols: support four layer load balancing including TCP protocol and UDP protocol, and seven layer load balancing including HTTP protocol and HTTPS protocol.

Certificate management: provide a centralized certificate management system for HTTPS protocol to meet users' requirements for reliable and secure transmission.
Multi level disaster recovery guarantee

Provide multi-level disaster recovery strategy and highly available user experience

Health check: regularly check the health of the backend ECS. Once an ECS exception is detected, the traffic will not be forwarded to the exception instance to ensure availability.

Cluster deployment: cluster deployment is adopted, and sessions between machines are synchronized. Machine failures and cluster maintenance are transparent to users, and there is no single point of failure.
Intelligent scheduling

Multiple scheduling algorithms and forwarding modes to improve application deployment flexibility

Scheduling algorithm: it supports weighted polling and minimum connection number scheduling algorithms, and can select corresponding algorithms to allocate user traffic according to their own needs.

Domain name URL forwarding: supports traffic scheduling based on different domain names and URLs to improve the flexibility of the application system.
Session Persistence

Forwards the access requests from the same user to the same server for processing within a certain time to achieve user access continuity

Session persistence based on source IP: Layer 4 (TCP/UDP) load balancing implements session persistence through IP addresses.

Cookie based session persistence: Seven layer (HTTP/HTTPS) load balancing supports session persistence through cookies.

Product Center

Convenient entrance

Security

other

contact us