Huawei Cloud Tech Essentials

Chapter 4:
Database and Data Governance

Huawei Cloud | Grow with Intelligence

 Objectives
Upon completion of this chapter, you will have:

⚫ Learned the databases of Huawei Cloud.

⚫ Learned database application scenarios.

⚫ Learned data governance.

Page 2
 Part 1:
Database

Huawei Cloud | Grow with Intelligence

Besides OBS and EVS, are there any
other services that can store data?

OBS EVS

Page 4
Different Data Types
Unstructured data Structured data
Book Name Published Year Author
The Great Gatsby 1925 F. Scott Fitzgerald

Lord of the Rings 1968 J. R. R. Tolkien

Document Package Audio/video

file Fully structured Semi-structured
{
Year: 2018,
Author: {
Byliner: J. R. R. Tolkien,
Drafter: Internet pal}
}

OBS EVS Relational Database Document

Service Database Service
RDS DDS
Page 5
Relational Database vs Non-Relational Database
Relational Non-Relational

Traditional applications require databases that can New Internet applications require databases that
Application model process transactions, complex queries, and can store massive data and handle a large number
associated queries, and maintain data integrity. of concurrent requests.

Within 10 ms
Milliseconds
Access performance Depending on query complexity and the amount of
The amount of stored data is irrelevant.
stored data

Data type Fully structured Semi-structured (key/value pairs and documents)

Vertical Vertical
Scaling mode
Horizontal: primary DB + read replicas Horizontal: cluster + read replicas

Depending on database configurations

Data consistency Strong consistency
Strong consistency or eventual consistency

Storage space limit Depending on the DB engine Depending on the number of deployed DB nodes

Page 6
Diversified Databases

No single service can meet all requirements.

Document Distributed GaussDB RDS for Relational GaussDB

Database Cache Service (for NoSQL) PostgreSQL Database (for MySQL)
Service Redis Service
DDS RDS

Page 7
Advantages of Cloud Native
Database optimization Database optimization Database optimization

Database development Database development Database development

Backup and restoration Backup and restoration Backup and restoration

High availability High availability High availability

Database engine management Database engine management Database engine management

Patch and configuration management Patch and configuration management Patch and configuration management

Software system Software system Software system

OS OS OS

Device device maintenance Device device maintenance Device device maintenance

Power supply, cooling, and cabinets Power supply, cooling, and cabinets Power supply, cooling, and cabinets

Buildings, locations Buildings, locations Buildings, locations

Cloud-native deployment
Traditional DC deployment Cloud host deployment

Green: Huawei Cloud's responsibilities Blue: Customer's responsibilities

Page 8
RDS Advantages

 A service that manages database engines and resources

 Compatible with mainstream relational database engines

◼ MySQL, PostgreSQL, SQL Server

 A cloud-native service that facilitates maintenance

 Built-in automated backup and patching

 HA configuration and automatic switchover/failover

Relational Database
Service within seconds

RDS  Read replicas

 Security certifications

Page 9
Features of Backup

02
Backups are Automated Binlog backups
stored in OBS Binlog backup can be restored
and can be and manual to a specific
restored to backup are point in time.
local disks. supported.

01 03

Page 10
Building a Simple Architecture Using a Database Service

• A database service, for example RDS,

functions as a database resource
Application Server manager.
App Server
• Applications access the database in the
Elastic Load original way.
RDS
Balance
ELB Application Server Database service • DB engines and versions are selected
App Server while tenants are creating databases.

However, this deployment architecture is

not safe.
Backups in
OBS

Page 11
Primary/Standby Deployment of RDS

• If the primary DB instance fails, workloads can be automatically switched to

the standby DB instance.
– The switchover duration varies depending on your DB engine. A switchover can be
performed within seconds.

• Primary and standby DB instances can be deployed in different AZs to enable

cross-AZ disaster recovery (DR).

• You can complete maintenance operations for the standby database and
then perform a switchover to ensure service continuity.

• Primary/standby deployment increases costs.

Page 12
HA Architecture Based on Primary/Standby Deployment

• This architecture eliminates the impact

RDS
App Server of a single point of failure (SPOF).
(Standby)
– High availability is achieved.
– Service continuity can be guaranteed
Elastic Load Balance
even if a node fails.
(ELB)
App Server

RDS • This is a typical way to use AZs.

(Primary) – The RDS service deployed in
primary/standby mode provides
complete functions in either of the AZs.

Backups in OBS
Page 13
Other Huawei Database Services

 Huawei DB engine

 Compatible with MySQL

 Several times higher performance than that of

native MySQL databases

 Strong backup and restoration capabilities

GaussDB(for MySQL) achieved with decoupled storage and compute

 High availability in cross-AZ deployment

Page 14
DDM Simplifies Database Use
Distributed Database Middleware (DDM) is a distributed database middleware service
that can work with RDS to horizontally expand database capacity so that regular users
can use distributed databases as easily as using a single-node database.

➢ SQL computing and transaction processing Service system

➢ Horizontal database scaling Still like using a single-node database

➢ Unlimited scaling of storage and compute

DDM
resources

➢ Single-node MySQL database experience RDS RDS

Page 15
What Will Happen If There Are a Large
Number of Concurrent Data Writes?

Page 16
GaussDB(for Mongo) Features

 Huawei-developed DB engine

 Compatible with MongoDB clients

 Compute resources scaled within minutes

and storage within seconds

GaussDB(for Mongo)  Cross-AZ HA

If necessary, you can select DDS

DDS instances that use the MongoDB engine.

Page 17
Features of Non-Relational Databases

01 03
There are no
logical Users can locate
relationships data using key
Each data record
between the data values rather
is a structured
records in a than through
document.
collection. complex queries.
02

Page 18
Precautions for Using Non-Relational Databases

• Define data keys and indexes for faster data location to avoid having to scan entire
collections.
– Do not use hot keys.

– Keep data records simple.

– Data can be located in milliseconds no matter how much data there is.

• If necessary, use optimistic locks for transactions in the application layer.

• Use clusters if there is too much traffic.

– Select an appropriate sharding algorithm to locate data.

– Distribute data access across all cluster nodes and avoid hot partitions.

Page 19
HA Architecture Using Non-Relational Databases

• GaussDB(for Mongo) is an
App server
alternative to RDS.
ELB – The type of databases depends on
GaussDB(for
App server Mongo) cluster the type of data and how it will be
accessed.

• GaussDB(for Mongo) clusters

Backups in OBS provide high availability.

Page 20
 Part 2:
Data Governance

Huawei Cloud | Grow with Intelligence

Data Processing Challenges

Structured and unstructured High data ingestion

Terabytes to OBS and database bandwidth
petabytes Logs, transactions, and Rapid data processing
clickstreams

Large volume Variety High velocity

Page 22
Typical Data Processing Method

Analysis and
Collection Visualization
processing

Data Insight

Storage

Page 23
Data Collection

RDS GaussDB(for OBS Distributed Data Ingestion

database Mongo) cluster Message Service Service
(DMS) (DIS)
for Kafka

Transaction data Files, logs, and raw data Stream data, IoT data, and
Structured transaction clickstreams
information

Page 24
Governing Data in Lakes
But is it really good to have so many choices
Challenges: for storing data?
1. Where is data stored?
2. What is the data format? Avoid data silos:
1. How to avoid repetition and conflict?
2. How to enable cross-component data
access?
3. How to transfer data from SaaS?
4. ...
Page 25
DataArts Studio Helps You Govern Data

 Starts from data governance planning.

 Data import

DataArts  One platform for data development and

Studio

governance

DataArts Studio  Data quality evaluation based on planning

 Data asset management and metadata

splitting

Page 26
Thank you. Bring digital to every person, home, and
organization for a fully connected,
intelligent world.

Copyright© 2024 Huawei Technologies Co., Ltd.

All Rights Reserved.

The information in this document may contain predictive

statements including, without limitation, statements regarding
the future financial and operating results, future product
portfolio, new technology, etc. There are a number of factors
that could cause actual results and developments to differ
materially from those expressed or implied in the predictive
statements. Therefore, such information is provided for reference
purpose only and constitutes neither an offer nor an acceptance.
Huawei may change the information at any time without notice.