Project Overview

As a leading securities firm in China, Haitong Securities embraced the development of financial technology actively and took the lead in launching the development of financial cloud for the industry. In 2017, the development process of Haitong Financial Cloud was further clarified. Combining its own business needs, Haitong Securities has established the goal of financial cloud development to achieve the transformation from traditional infrastructure to hybrid cloud architecture.

• Based on the OpenStack framework, Haitong Securities has built an IaaS to SaaS financial cloud platform with certain service capabilities and financial architecture characteristics. From a multi-cloud perspective, it provides users with a fully autonomous operation experience and solves the problem of unified asset management, distribution, scheduling… etc.

  
  

• The platform has thousands of tenants, all users were facing problem with data mobility. Therefore, it is considered to build in a cloud-native migration solution to solve the problem, enable users to migrate the business systems across different resource pools flexibly.

  
  

  Solution

Integrated in the form of REST API, providing self-service migration module

All the migration service functions are provided to the cloud management platform in the form of REST API, so that users do not need to operate the migration platform separately, users can complete migration on the cloud management platform through simple configuration without downtime.

Achieve multi-cloud auto-adaptation through cloud platform API calls

Complete multi-cloud adaptation through the cloud platform interface driver layer, users can migrate the systems to specific cloud resource pool without restriction on cloud platform type, storage type… etc., enabling system resource mobility.

Minimalist wizard interface

Complete virtual machine migration through six visualized steps: 

Basic information collection → Pre-check → Registration → Synchronization → Startup → Cleanup 

Customer Outcomes

Enterprise cloud platform integrates migration function, empowering platform users

Enterprise leverages migration service as platform capability to empower its tenants, reducing the complexity and difficulty of operations and management, becoming the core advantage of enterprise cloud management platform.

Self-service migration, enhancing data mobility

Solved the problem of data resource mobility for multiple users, users do not need to manage cloud platform resources throughout the migration process, all resources are created automatically, users can select any data resources for migration.

Migration Scenario

• Source platform: VMware virtual machine * 2, physical machine (X86) * 2

• Transmission: 50MB/S

• Target cloud platform: FIT2CLOUD CloudExplorer CMP

• Switching window: None, tenants can set their own schedule for business migration

  
  

Migration Resources

• Tool： HyperMotion

• Manpower: 1-2 person on site (customer operations & maintenance), 1 person remote support (migration vendor)

• Efficiency: 800 tenants on the platform shared one account for migration, saving time and resources

  
  

Migration Systems

Online trading system, quantitative trading system, options trading system, financial trading system, OA system, company website system.

Migration Process

Pre-migration

Research：

Collect basic system information → System dependency analysis → Review migration list → Review migration SLA and resource configuration → Plan for business cutover operations

During migration

• Environment deployment: Resource preparation for migration deployment → Migration software installation → Network configuration → Source platform configuration → Target cloud platform configuration

• Migration implementation: Source platform data collection → Data synchronization → Host boot up on cloud platform 

• Validation (preliminary): Confirm host status → System configuration →Validate business service

Post-migration

Service cutover: Pre-cutover preparation → cutover operation → service validation