I. Introduction

In the digital age, enterprises are faced with more and more database and application migration requirements. To meet this demand, HUAWEI CLOUD has launched a new database and application migration solution (database and application migration UGO). This solution is mainly used to solve the difficult problems of enterprise data migration (such as syntax incompatibility, technical support, high transformation cost...especially for heterogeneous databases, which cannot be migrated, moved, and costly). It can help users easily migrate source databases to target platforms, reduce migration costs, and increase conversion rates.

2. Database and application migration UGO definition

1. Definition of UGO

Database and Application Migration UGO (Database and Application Migration UGO, hereinafter referred to as UGO) is a professional service focusing on the migration of heterogeneous database structures .

Through UGO, DDL, DML, and DCL in the source database can be automatically converted into the SQL syntax of HUAWEI CLOUD GaussDB/RDS with one click. Through the two core functions of database evaluation and object migration and automatic syntax conversion, possible transformation work can be identified in advance, the conversion rate can be improved, and the cost of user database migration can be minimized.

2. Simplified diagram of heterogeneous database migration

Description: It is mainly divided into three parts, source database, migration tool (service) and target database. Source databases mainly include offline databases (such as oracle, mysql...) and cloud databases (such as RDS, self-developed...); database migration tools mainly include UGO (grammar migration), DRS (data migration and verification); target databases are mainly Huawei cloud databases (such as GaussDB, RDS, etc.).

3. Pain points of database migration

Some pain points of database migration:

High complexity: Traditional database migration usually requires extensive modifications and adjustments to the source database, including data structures, storage methods, and indexes. These modifications may cause problems such as application incompatibility or performance degradation, thereby increasing the complexity and risk of migration.
High cost: Traditional database migration usually requires a professional team to plan, design and implement, requiring a lot of manpower, material resources and time. At the same time, if something goes wrong during the migration, additional repairs and adjustments are required, increasing costs and risks.
High risk: Traditional database migration involves interactions and dependencies between multiple systems and applications. Once a problem occurs, it may affect the stability and availability of the entire system. In addition, due to differences between the source and target databases, issues such as data loss, corruption, etc. may occur during the migration process, resulting in business interruption and loss.
Poor scalability: Traditional database migrations are usually one-time events that cannot be scaled and upgraded once completed. If you need to add new features or support more users, you need to redo the database migration, which will increase the cost and risk.

There are a series of pain points in database migration. These problems will not only affect the operational efficiency and competitiveness of the enterprise, but also increase the cost and risk of the enterprise. Therefore, adopting new database migration tools and technologies is the key to solving enterprise database migration.

4. Introduction of database and application migration UGO capabilities

UGO is a specialized tool focusing on heterogeneous database object migration and application migration. Through the two core functions of pre-migration evaluation and structure migration , it helps users easily realize one-click cloud access and one-click database switching.

Source database portrait: Algorithm training is carried out with the business scenarios of massive source databases as samples and key database indicators as characteristics to obtain the portrait information of the source database instance.
Type selection and specification: Using the image of the source library as input, through the UGO target recommendation model, recommend a suitable combination of target library and specification for the user. The selection scheme of the target library will dynamically generate different recommended combinations according to the image input of the source library.
Compatibility analysis: take the image of the source library as input, and analyze the compatibility of 14 core object types through the conversion rate of the UGO kernel to the target library.
Workload evaluation: With the source database portrait as the input and the human migration cost of massive offline scenarios as the evaluation baseline, the UGO evaluation model can be used to analyze the migration cost of the database instance, including the workload of DBA, development, and testing.
Structural Migration: With pre-migration evaluation as input and plan guidance, users are supported to customize and filter migration objects before conversion. After the conversion is completed, mark the conversion failed objects and failure reasons. Users can correct objects according to the failure reasons. Until all objects are verified successfully.
SQL migration: It is divided into two parts. The first part is to deploy Agent on the application side to capture traffic, collect SQL statements submitted from the application side to the database, and perform dynamic conversion through the UGO kernel. The converted SQL is played back on the target side and generates performance analysis reports. The second part collects the SQL submitted to the source database by the application side, converts it through the UGO kernel after collection, verifies the conversion accuracy in the target database, and generates a SQL translation report for developers to modify the application program.

5. Introduction to UGO Scheme of Database and Application Migration

With the cloud-based development of GaussDB, database and application migration can be performed easily and conveniently through UGO. The migration process of UGO is mainly divided into five steps:

The first step: the source database, which supports offline databases or other cloud databases as the source.
Step 2: Object collection, which collects data from the database. In addition to object structure and SQL, the collected information also includes information such as capacity, performance, complexity, and configuration.
Step 3: Migration assessment, analyzing the source database for portraits, evaluating the compatibility, selection, specifications, workload, risks, etc. of the target database, and predicting risks in advance.
The fourth step: conversion verification, equivalent grammar conversion is performed on the collected objects, and a verification test is performed on the converted grammar.
Step 5: Migrate online (target library), and automatically apply objects that pass the verification to the target library.

For best practices, please refer to the official website "Oracle Structure Migration to GaussDB":

Solution Overview_Database and Application Migration UGO_Best Practice_Oracle Structure Migration to GaussDB_HUAWEI CLOUD

6. Summary

The essence of the database object migration process is to execute the SQL statement of the database object converted by UGO in the target database. After the database migration phase is completed, the object attributes of the source library and the target library are extracted for comparison to ensure that there is no object loss after using UGO for database object migration.

UGO is responsible for risk and workload assessment of the source database, and automatically converts heterogeneous database syntax. It is understood that 95% of the automation has been achieved so far; DRS is responsible for real-time online migration and synchronization of full and incremental data, and ensures zero data loss through data consistency comparison. Huawei Cloud Database provides "GaussDB + UGO + DRS" one-stop migration solution to the cloud, which can help users identify migration risks in advance, improve migration efficiency, and minimize users' database migration costs.

--Finish.

GaussDB cloud database supporting tool UGO