Load Balancer

Load balancer is a load balancing service that distributes traffic to multiple cloud servers. Automatically distribute the application's external service capabilities across multiple instances, increasing the availability of your application by eliminating single points of failure, enabling you to achieve higher levels of application fault tolerance, providing seamlessly the load required to distribute application traffic capacity, providing you with efficient, stable and secure services.

Order Now Console

Support 5 Unconditional days refund service


  • Exclusive Instance Performance
    Single instance exclusive computing resources, public network IP independent, the IP number is not exceedable
  • Support Multiple Protocols
    Supports four layers (TCP/UDP) and seven layers (HTTP/HTTPS) protocol load to meet different application scenarios.
  • Improve Cloud Server Availability
    Improve front-end access to two or more business nodes on the back end, improving cloud server availability
  • Easy to Scale Out
    The backend node can freely increase or decrease according to the actual scene requirements.
  • Suitable for High Concurrent Access Services
    High concurrent service access is balanced by multiple nodes, or with different weights according to the specified weight value
  • Eliminate Single Node Failures
    Use multiple nodes to provide business access processing, thus avoiding the problem of single node failure causing business stagnation


  • Advantages
  • Performance
  • Availability
  • Scalability
  • Costs
  • Security
  • ZKEYS Load Balancer
  • The maximum number of concurrency in a single cluster exceeds 10 million, and it can handle peak traffic of 20 Gbps. You can confidently deploy e-commerce websites, social APP software, and game services with over 10 million daily visits.
  • High availability, providing redundant backup, automatic failover, ensuring that the backend provides services normally, the cluster system will automatically reject the fault instance, and the user only needs to consider the backend service code.
  • The cluster scales horizontally according to the service load and expands the external service capability of the application system, so that you do not need to worry about the expansion problem. The user accessing the resources in this process is transparent.
  • No need to purchase load balancing hardware repeatly, which greatly saves load balancing hardware overhead and saves up to 90%.
  • Through professional DDoS protection equipment, users can provide fine defense against DDOS attacks for Internet applications, such as UDP, Flood attacks, SYN Flood attacks and CC attacks.
  • Traditional Load Balancing
  • Traditional hardware devices such as F5 BIG-IP and Netcaler support only a million connections, and the throughput is less than 5 Gbps, which cannot meet the sudden and massive Web access requests.
  • The dual-machine HA solution provided by traditional hardware devices is not enough to cope with the complicated and varied public network environment.
  • When traditional hardware devices, such as F5 BIG-IP and Netscaler, encounter service bottlenecks, they need to re-purchase higher-standard hardware devices for redeployment and affect business development.
  • Traditional hardware devices cost tens of thousands to hundreds of thousands per 10,000, making deployment difficult; requiring professional operation and maintenance personnel, hardware costs and labor costs have increased dramatically.
  • Traditional hardware devices have limited DDoS intrusion capabilities
  • Open Source Solution
  • Self-built Nginx, HAProxy and other programs, the maximum number of connections is often only tens of thousands, unable to cope with the massive concurrent requests of the Internet.
  • Strict technical requirements, the scheme is cumbersome, and the availability cannot be guaranteed.
  • Manual intervention is required, and it is time consuming, which has a great impact on the business and risks of operational errors.
  • Open source solutions are free, but there is no uniform standard and high maintenance costs
  • Lack of mature security solutions.


Multiple Protocol Support
Multi-protocol support to meet your diverse application scenarios
Specific protocol: Supports four layers of load balancing including TCP and UDP, and seven layers of load balancing including HTTP and HTTPS.
Certificate Management: Provides a centralized certificate management system for the HTTPS protocol to meet the needs of users for reliable and secure transmission.
Multi-level Disaster Recovery
Provide multiple levels of disaster recovery strategies and highly available user experience
Health Check: Checks the running status of the back-end cloud server periodically. Once the cloud server is detected abnormally, the traffic will not be forwarded to the abnormal instance to ensure availability.
Cluster Deployment: Cluster deployment, session synchronization between machines, machine failure and cluster maintenance are transparent to users, with no single point of failure.
Intelligent Scheduling
Multiple scheduling algorithms and forwarding modes to enhance application deployment flexibility
Scheduling Algorithm: Supports weighted polling and minimum connection number scheduling algorithm, and can select the corresponding algorithm to allocate user traffic according to its own needs.
Domain Name URL Forwarding: Supports traffic scheduling based on different domain names and URLs to improve application system flexibility.
Session Retention
Forward access requests from the same user to the same server for a certain period of time to achieve user access continuity
Source IP-based Session Persistence: Layer 4 (TCP/UDP) load balancing implements session persistence through IP addresses.
Cookie-Based Session Persistence: Layer 7 (HTTP/HTTPS) load balancing supports session persistence through cookies.

Architecture Deployment