• The PC era is essentially in which computers are networked, and personal

computers are connected through servers. Now, in the mobile era, we can assess
the Internet through mobile phones. With the advent of 5G, all computers,
mobile phones, and intelligent terminals can be connected, and we can enter an
era of Internet of Everything (IoE).
• In the IoE era, the entire industry will compete for ecosystem. From the PC era to
the mobile era, and to the IoE era, the ecosystem experiences fast changes at the
beginning, then tends to relatively stable, and rarely changed when it is stable. In
the PC era, a large number of applications run on Windows, Intel chips, and x86
architecture. Then, browsers come with the Internet. In the mobile era,
applications run on iOS and Android systems that use the ARM architecture.

• Compared with the previous generation, the number of devices and the market
scale of each generation increase greatly, presenting future opportunity. As the
Intel and Microsoft in the PC era and the ARM and Google in the mobile era,
each Internet generation has its leading enterprises who master the industry
chain. In the future, those who have a good command of core chips and
operating systems will dominate the industry.
• The Internet brings a large amount of traffic, users, and data, so enterprises need
to continually purchase traditional IT devices to keep pace with their rapid
development. Therefore, the disadvantages of traditional IT devices gradually
emerge.

▫ Long procurement period causes slow rollout of new business systems.

▫ The centralized architecture has poor scalability and can only increase the
processing performance of a single node.

▫ Traditional hardware devices exist independently, and their reliability

depends only on software.

▫ Devices and vendors are heterogeneous and hard to manage.

▫ The performance of a single device is limited.

▫ The utilization of devices is low, while the total cost remains high.
• Let's discuss

▫ How to solve these pain points? Think over advantages of cloud computing
that can solve these pain points, so you can have a better understanding of
cloud computing.
• The traditional IT architecture consists of hardware and software, including
infrastructure, data centers, servers, network hardware, desktop computers, and
enterprise application software solutions. This architecture requires more power,
physical space, and capital, and is usually installed locally for enterprises or
private use.
• With the virtualization technology, computer components run on the
virtualization environment, not on the physical environment. Virtualization
enables maximum utilization of the physical hardware and simplifies software
reconfiguration.

• With cloud transformation, enterprise data centers are transformed from

resource silos to resource pooling, from centralized architecture to distributed
architecture, from dedicated hardware to software-defined storage (SDS) mode,
from manual handling to self-service and automatic service, and from distributed
statistics to unified metering.
• Cloud computing has the following characteristics:

▫ Broad network access

▫ Fast and elastic scaling

▫ On-demand self-service

▫ Resource pooling

▫ Metered services
• What are the data sources of cloud computing in daily life?

▫ Cloud album, such as Baidu Cloud and iCloud Shared Album

▫ Cloud music, such as NetEase Cloud Music, Kugou Music, Kuwo Music, and
Xiami Music

▫ Cloud video, such as Baidu Cloud and Tencent Cloud Video

▫ Cloud documents, such as Youdao Note, and Shimo document

• These apps are based on cloud computing, making our lives more convenient.
Cloud computing allows enterprises to provide better products to attract more
users.
• Driven by the requirements of the government, transportation, electric power,
medical care, education, finance, and military industries and enterprises, the
video conferencing market in China has an average annual growth beyond 20%.
Currently, only less than 5% of enterprises in China have video conference rooms,
and more and more enterprises are aware of the importance of efficient
collaboration. Therefore, the video conferencing system has become
indispensable for efficient office work.

• HUAWEI CLOUD Meeting can be used by enterprise office, telemedicine, smart

education, and enterprise organization construction.
• Online services are the most typical application of e-Gov and are used in multiple
cities. With online services, applicants fill in the application information and
submit supporting documents. The administrative approval center approves
applications by streamlining cross-department data. By moving services to the
cloud, governments can greatly reduce expenditure, cloud service providers can
gain new development benefits, and citizens can acquire services more and more
conveniently.

▫ Guidance: All policies, bulletins, and processes are released through

information guidance, making service handling processes clear for citizens
and enterprises. Intelligent Process Automation (IPA) robots are provided to
guide users.

▫ Handling: Based on big data and AI technologies, fields in documents can

be auto-populated.

▫ Review: AI technologies pre-review documents, improve review efficiency

and quality, and reduce pressure on staff. Real-Time Communication (RTC)
audio and video technologies can implement contactless online pre-review.

▫ Collaboration: RPA technology handles all work items through the

workbench, effectively collaborating with functional agencies. All
application and service entries can be managed in a unified manner.
• Public cloud is the main form of cloud computing, which is developing
considerably in China. Public cloud vendors can be classified as follows:

▫ Traditional telecom infrastructure carriers, including China Mobile, China

Unicom, and China Telecom

▫ Local government cloud computing platforms

▫ Public cloud platforms built by Internet giants, such as Alibaba Cloud and
Tencent Cloud

▫ Some IDC carriers, such as 21Vianet Group

▫ Enterprises with foreign technical background or introducing foreign cloud

computing technologies, such as Fengqi.Asian Cloud
• Private cloud is a cloud infrastructure operated solely for a single organization.
All data of the private cloud is kept within the organization's data center.
Attempts to access such data will be controlled by ingress firewalls deployed for
the data center, offering maximum data protection.

• Public cloud service provider owns and operates the cloud infrastructure and
provides cloud services open to the public or enterprise customers. This model
gives users access to convenient, on-demand IT services, comparable to how they
would access utilities like water and electricity.

• A hybrid cloud is a combination of a public cloud and a private cloud or on-

premises resources, that remain distinct entities but are bound together, offering
the benefits of multiple deployment models. Users can migrate workloads across
these cloud environments as needed.
• Infrastructure as a Service (IaaS): The cloud platform provides infrastructure
(such as servers, storage devices, networks, and virtual resources) and maintains
related resources. Users only need to pay attention to the system and application
layers.

• Platform as a Service (PaaS): The cloud platform provides infrastructure (such as

servers, storage devices, networks, and virtual resources) and application
deployment environment (such as the operating system, middleware, and
software running environment) and maintains related resources. Users only need
to focus on applications and data.

• Software as a Service (SaaS): The cloud platform provides all resources, services,
and maintenance. Users only need to use applications.

• Compared with the conventional IT entire-process and all-device procurement

mode, the cloud service-oriented mode provides IT devices as services that allow
customers to select on demand, which has more advantages in flexibility, and
low cost.
• Cloud computing integrates hardware resources into one in software, and
dynamically allocates the resources to applications in software, which greatly
improves the resource usage. Cloud computing also supports auto scaling, and
optimizes the work efficiency. By building high-specification cloud data centers
and introducing the automatic scheduling technology, data storage is more
centralized, and data assets are more effectively used, achieving energy saving,
emission reduction, and easier maintenance. In this way, lower costs and higher
efficiency are achieved in each dimension.
• Cloud computing has five main benefits.
▫ On-demand self-service: Customers can deploy processing services based on
actual requirements on the server running time, network, and storage, and
do not need to communicate with each service provider.
▫ Broad network access: Various capabilities can be obtained over the
Internet, and the Internet can be accessed in standard mode from various
clients, such as mobile phones, laptops, and PADs.
▫ Resource pooling: Computing resources of the service provider are
centralized so that customers can rent services. In addition, different
physical and virtual resources can be dynamically allocated and reallocated
based on the customer requirements. Customers generally cannot control
or know the exact location of the resources. The resources include the
storage devices, processors, memory, network bandwidth, and virtual
machines.
▫ Quick deployment and auto scaling: Cloud computing can rapidly and
elastically provide computing capabilities. A customer can rent unlimited
resources and purchase required resources at any time.
▫ Metered services: Cloud services are billed based on the actual resource
usage, such as the CPU, memory, storage capacity, and the bandwidth
consumption of cloud servers. Cloud services provide two billing modes: pay-
per-use and yearly/monthly.
• Massive scale: Cloud computing service is in large scale as it centralizes IT
resource supply. This makes cloud computing different from conventional IT.
• Homogeneity: Homogeneity can also be understood as standardization, which is
similar to power utilization. Voltage and socket interface should be the same for
various electrical appliances and devices.
• Virtualization: Virtualization has two meanings. One is accurate computing units.
If a cake is too large for one person, it is better to divide it into small pieces to
share. That is, with smaller computing units, IT resources can be fully used. The
other meaning is the separation of software and hardware. Before virtualization,
software and specified hardware are bound together, and after virtualization,
software can be freely migrated on all hardware, which is like renting a house
instead of buying one.
• Elastic computing: Elastic computing means that IT resources can be elastically
provided.
• Low-cost software: Low-cost software is provided to meet the market
competition and requirements. Cloud computing, with low individual technical
skill and financial requirements, makes IT easy to use. Small and micro startups
are always willing to enjoy the more IT services at the lowest cost. Based on this
situation, low-cost software is required to earn money at small profits but quick
turnover.
• Geographic distribution: As the broad access mentioned above, IT services can be
provided anytime and anywhere. From the perspective of users, cloud computing
data centers, are geographically distributed and the performance of network
bandwidth varies by regions. Large public cloud service providers have dozens or
even hundreds of data centers or service nodes to provide cloud computing
services to global customers.
• Compute services provide computing power required for running services such as
websites, office software, and data analysis. Currently, typical compute cloud
services are VMs and containers.
• Network services provide resource connectivity and isolation, such as data center
networks and campus networks. On the cloud, VMs use virtual networks (for
example, VPC) that have the logical topology similar to that of traditional
networks.

• Storage services include:

▫ Block storage: features high performance and low latency, meeting

different high I/O service requirements.

▫ File storage: allows file sharing among multiple servers or enterprise

departments.

▫ Object storage: features a flat, easy scale-out architecture, which is suitable

for cloud storage. It is mainly used for massive data storage, cold data
backup, and software repository.
• In computer technologies, virtualization is a resource management technology. It
abstracts various physical resources of a computer, such as CPU, memory, disk
space, and network adapters, converts the resources, and presents the resources
for segmentation and combination into one or more computer configuration
environments. In this way, the uncut barriers between physical structures are
broken, allowing users to use computer hardware resources in a better way than
the original configuration.

• As shown in the figure, a physical server is divided into multiple files through
virtualization, and each file represents a VM.
• Virtualization is a key technology of cloud computing. It aims to abstract physical
resources into logical resources for flexible allocation, Virtualization offers
scalable, distributed, and HA resources for cloud computing.
• Cloud computing allows users to use cloud resources on demand, relying on the
virtualization technology.
• Features of VMs:

▫ Partitioning: Multiple VMs can run on one physical server, which means
that the virtualization layer can allocate resources of a physical server to
multiple VMs. This is called partitioning.

▫ Isolation: If one VM on a server is faulty or infected with viruses, the other

VMs can still run properly.

▫ Encapsulation: VMs exist in the virtualization system as files. You can

migrate VMs by cutting/copying and pasting files.

▫ Independence: After being migrated to another physical server, a VM can

properly run without any modification on the server because VM OSs are
decoupled from physical hardware.
• Hypervisor: It is also called virtualization software or VM monitor. Hypervisor is
used to create and run VMs on physical servers. The mainstream open-source
virtualization technologies are Xen and KVM.
• A computer system consists of CPU, memory, disk, and network resources.
Compute resources include CPU, GPU, and memory.

• Central Processing Unit (CPU) is the computing and control core of a computer
system, which processes information and executes programs.

• Memory is an important component of a computer system. It is used to store

CPU computing data and exchange data between memory and external storage
(such as hard disks).

• Graphics Processing Unit (GPU) is a microprocessor that performs image

computation on PCs, workstations, game consoles, and mobile terminal devices
such as tablet and smartphones.
• An Elastic Cloud Server (ECS) is a VM on the cloud, consisting of vCPUs, memory,
OS, and EVS disks. After buying an ECS, you can use it on the cloud just like you
would use your local PC or physical server.
• Auto Scaling (AS) automatically scales compute resources based on your
demands and the AS policies you have configured, properly adjusting the number
of ECSs as the service load changes over time.

• An image is a template used to create servers or disks. Image Management

Service (IMS) provides image lifecycle management. With the IMS, you can
create a system or data disk image from a server or an external image file, or
create a full-ECS image from an ECS or a backup of an ECS.
• Containers can:

▫ Package software into standardized units for development, migration, and

deployment.

▫ Isolate compute, storage, network, and other resources.

▫ Start, stop, deploy, and migrate applications agilely and instantly.

▫ Allow developers to focus on R&D and O&M engineers to focus on system

maintenance.
• Container technology was born in 1979, introduced as the chroot operation in
UNIX. Chroot provided an isolated file system for each progress so their root
directories can be easily changed. This is the origin of OS virtualization.
• In 2000, BSD released FreeBSD Jails based on chroot. In addition to file system
isolation, FreeBSD Jails isolate users, networks, and other resources. An IP address
was assigned to each jail, which is an independent, smaller computer system, for
independent software installation and configuration.
• In 2005, SWsoft released OpenVZ, which was similar to Solaris Containers.
OpenVZ uses a modified Linux kernel to provide virtualization, isolation, resource
management, and checkpoints. Since then, kernel virtualization has become a
mainstream solution.
• In 2006, Google launched Process Containers. Process Containers, renamed as
control groups (cgroups) later, were designed for limiting, accounting, and
isolating resource usage (CPU, memory, disk I/O, network) of a collection of
processes. In 2007, cgroups were merged into Linux kernel 2.6.24.
• In 2008, LXC (the first, most complete implementation of Linux container
manager) was implemented using cgroups and Linux namespaces. LXC can work
on a single vanilla Linux kernel without requiring any patches.
• Containers and VMs have similar advantages in resource isolation and allocation
but different functions, because containers virtualize OSs instead of hardware.
Containers are more portable and efficient.
• There is no virtualization layer in the container architecture. Therefore,
containerization is called lightweight virtualization. Applications running in
containers have better performance than those in VMs.
• Containers have become popular because of many benefits, including:
▫ Agile building and deployment of applications: The creation of container
images is easier and more efficient than that of VM images.
▫ Consistent development, integration, and deployment: Containers can be
quickly restored using images. You can customize new images for consistent
container building and deployment.
▫ Portability across clouds and OSs: Containers can run on Ubuntu, RHEL,
CoreOS, Google Kubernetes Engine, physical servers, etc.
▫ Application-centered management: The abstraction is improved from
virtualizing hardware for OS isolation to virtualizing an OS for application
isolation.
▫ Loosely coupled, distributed, elastic, independent miscroservices:
Applications are divided into independent, small units and can be deployed
and managed separately instead of running on a single large server.
▫ Isolated resources: Application performance can be predicted.
▫ High resource utilization: Resources can be fully used.
• Containers are an abstraction at the application layer. A container packages up
code and its dependencies required for proper running of an application. Multiple
containers can run on the same server with a shared OS kernel. Each container
runs as an independent process in the user space. Containers take up less space
than VMs, process more applications, and require less CPU and memory.
• Virtual Machines (VMs) are an abstraction of physical hardware and turn one
server into multiple servers. The hypervisor allows multiple VMs to run on the
same physical server. Each VM has its own OS, applications, necessary binaries,
and libraries, taking up tens of GB. The startup speed of a VM may be slow.

• Container image: A container image is dedicated to running a specific service and

usually contains only the resources required for running the service. Many widely
used images are tens of MB or less in size.

• VM image: A VM image offers the operating environment (including the OS

kernel) required by common processes and provides a complete collection of
functions. The minimum size of a VM image is hundreds of MB.
• Cloud Container Engine (CCE) is a high-performance, high-reliability service
through which enterprises can manage containerized applications. CCE supports
native Kubernetes applications and tools, allowing users to easily set up a
container runtime environment on the cloud. CCE Turbo clusters run on the cloud
native 2.0 infrastructure, accelerating compute, network, and scheduling.
• Cloud Container Instance (CCI) is a serverless container engine that allows users
to run containers without creating or managing server clusters.

• SoftWare Repository for Container (SWR) allows users to easily manage the full
lifecycle of container images and facilitates secure deployment of images for your
applications. Users can upload, download, and manage container images through
SWR Console, community CLI, or SWR APIs.

• SWR can either work with CCE and CCI or be used as an independent container
image repository.
• Networks bridge devices and VMs and allow them to communicate with each
other. Therefore, networks are essential for ICT infrastructure.
• Broadcast and unicast: The communication between two devices is like that
between people. The unicast, like one person talking to another, refers to that
the information is sent and received between two nodes. The broadcast, like one
person using a loudspeaker to talk to many people, has higher communication
efficiency and ensures that the information can be sent to all related devices.
• Router: A router is a hardware device that connects two or more networks. It
works as a gateway to read the address of each data packet and decide how to
forward it.

• Default gateway: To understand the default gateway, we need to know what a

gateway is. A gateway is a device that connects a subnet to an external network.
When a device sends information to a host, a subnet mask determines whether
the destination host is on the local subnet according to the destination address. If
the host is on the local subnet, the device can directly send information to the
host. If not, the device will first send the information to the default gateway or
router, which then forwards the information to other networks to reach the host.

• Virtual Local Area Network (VLAN): VLAN is a group of logical devices and users,
which are organized based on functions, departments, and applications,
regardless of their physical locations. Such devices and users communicate with
each other as if they are on the same network segment. VLANs can be used to
isolate different services.
• Our PCs can access the Internet through a router. Likewise, servers can be
connected to the Internet by using a router.

• A router is a gateway device that operates on the third layer of the OSI Model,
the network layer. It stores and forwards packets between different networks,
and routes data from one subnet to another. In network communications, routers
can determine network addresses and select IP routes. Routers can flexibly set up
connections for networks and send packets between them through different
media access control mechanisms. Routers accept information only from the
source and other related routers, functioning as interconnection devices on the
network layer.
• A network switch is used to forward electrical signals, and establishes an
exclusive electrical signal route for any two nodes connected to the switch.
Ethernet switches are most commonly used. Other common switches include
telephone voice switches and fiber switches. Switching allows devices to
automatically or allows you to manually send information to an appropriate
route, meeting the requirements of both communications ends. A switch has
multiple ports, with each port providing the bridging function. A port can be
connected to a local area network (LAN) or a high-performance server or
workstation.
• On a conventional network, Layer 2 switches use VLANs to isolate network
planes.
• For safety and management purposes, a large local area network (LAN) must be
divided into several small LANs to reduce the impact of broadcast storms, so the
virtual local area network (VLAN) technology is widely used. Communications
between different VLANs are forwarded by routers. With the increase of access
across networks, if only routers are used, the network scale and access speed are
restricted because there are limited port quantity and the routing speed is slow.
To address this, Layer 3 switches are developed. Layer 3 switches are designed for
IP addresses. These switches provide simple APIs and are strong in processing
Layer 2 packets, suitable for routing and switching data in large LANs. Layer 3
switches not only replace or partially complete the function of traditional routers
in the third layer of the network model, but also have almost the same switching
speed as the second layer. And the price of Layer 3 switches is cheaper.
• The onboard NIC provides network expansion capabilities. It transmits data from
servers to other devices, providing application services externally.

• Commonly supported NIC speed rates include 100 Mbit/s, 1 Gbit/s, and 10 Gbit/s.
• Why is a virtual network required?

▫ VMs hosted on a physical machine may be in different IP address ranges, so

these IP address ranges need to be isolated. In addition, VMs need to share
the same physical NIC to access external networks. Therefore, virtual
switches are used on servers to construct virtual networks.
• In network virtualization, the first problem to be solved is how to map virtual
NICs of the VMs to the physical NICs of the physical server where the VMs are
hosted. As shown in the figure, we can use network bridges, NAT and virtual
switches to solve this problem.
• Both a bridge and NAT can forward the traffic of different VMs to physical NICs
so that data packets can be routed from the server to the physical switch,
implementing the communication between VMs and between VMs and external
networks.

• Virtual switches also have the bridging function. A virtual switch has a table that
defines mapping between MAC addresses and ports to isolate collision domains.
Simply speaking, a bridge connects different physical LANs at the data link layer.

• NAT forwards the traffic to external networks through translating network

addresses. NAT not only avoids the lack of IP addresses, but also protects
computers on the private network from being attacked by other networks.
• Open vSwitch (OVS): An OVS is a software-based open source virtual switch. It
supports multiple standards and protocols with additional support for the
OpenFlow protocol, and can be integrated with multiple open-source
virtualization platforms. An OVS can be used to transmit traffic between VMs
and implement communication between VMs and external networks.
• Enhance vSwitch (EVS): An EVS is an enhanced OpenFlow-compliant virtual
switch that improves the I/O performance based on the OVS forwarding
technology. I/O performance is significantly improved by using the Intel DPDK
technology and using user-mode processes rather than NICs to send and receive
data.

• On an OVS, data is received and sent in the kernel mode, but on an EVS, data is
processed in the user mode.

• Distributed Virtual Switch (DVS): Same as a physical switch does, a DVS

constructs the network between VMs and connects VMs to external networks.

• A virtual NIC of a VM communicates with an external network by connecting to

the DVS, then by connecting to the physical NIC of the host through the DVS
uplink.

• Compared with traditional switches, using virtual switches reduces network

devices and simplifies the network architecture, relieving the pressure of system
management and maintenance.
• A Virtual Private Cloud (VPC) is a private and isolated virtual network on
HUAWEI CLOUD. Users can configure IP address ranges, subnets, and security
groups, assign EIPs, and allocate bandwidths in a VPC.
• Public NAT gateways and private NAT gateways are used in different scenarios to
provide the network address translation. A public NAT gateway provides SNAT
and DNAT so that cloud servers in a VPC can share EIPs to access the Internet. A
private NAT gateway provides the network address translation for servers in a
VPC.
• The EIP service provides independent public IP addresses and bandwidth for
Internet access. EIPs can be bound to or unbound from ECSs, BMSs, virtual IP
addresses, load balancers, and NAT gateways. This service provides various billing
modes to meet diverse service requirements, and allows cloud servers in a VPC to
share the same private IP address to access or provide services accessible from an
on-premises data center or a remote VPC.
• How does storage media work?

A storage medium is any technology -- including devices and materials -- used to

place, keep and retrieve electronic data.
• In terms of data storage, the existing cloud storage products can achieve higher
efficiency at lower cost. Therefore, cloud storage will be an inevitable choice for
individuals and enterprises.
• Traditional servers have computing and storage coupled, and use their local
physical disks to store data. This is what we call the traditional block storage,
where a disk is connected to a server through a bus, delivering low latency.
However, the number of disks attached to the server is limited, so traditional
servers have poor performance in capacity, bandwidth, and reliability. The
explosive data growth poses high requirements on data reliability, which requires
decoupled compute and storage. To address this, storage arrays appear.
Traditional disk arrays comprise controllers and disk enclosures. Two or more
controllers can be used to provide high reliability. By adding disk enclosures, the
capacity of disk arrays can be hundreds of thousands of times larger than that of
local disks. Disk arrays independently connect to servers through FC switches or
IP switches. This is today's block storage.

• As the IT system further develops, enterprises want their files to be shared

among multiple hosts for concurrent access. This is shared file storage. Shared
file storage shares data in the same data center or equipment room.

• As more and more Internet applications need to access data over the Internet
using terminal devices, object storage that supports HTTP and HTTPS protocols is
widely used. Object storage allows applications to access data by calling APIs and
adopts a distributed architecture featuring large capacity and high reliability.
• Block storage cannot be directly used in an operating system. Before using a
block device, you must format it and create a file system on it. Data in the
operating system is stored as files.
• Block storage has the lowest latency among the three types of storage and is
ideal for mission-critical applications such as databases and ERP systems.
• Network File System (NFS): NFS is a file sharing protocol between UNIX
operating systems. It commonly applies to Linux clients.
• Common Internet File System (CIFS): CIFS is a protocol that allows programs to
access files on remote computers over Internet. It mainly applies to Windows
clients.
• File storage provides PB-level capacity and ms-level latency and is perfect for
scenarios where data needs to be shared among multiple compute nodes , such
as HPC and office automation.
• Object storage has large-scale data management capability, which is its biggest
advantage over file storage. File Storage uses a hierarchical structure to manage
all files and directories. If there are too many files or directories stored, the
search performance will be greatly reduced. Object storage provides a flat
structure where all objects are stored at the same logical layer. This keeps the
object search speed almost unchanged even if there are tens of billions of objects.
However, object storage uses application-level APIs instead of system-level APIs.
Traditional applications need to be redeveloped when being migrated to object
storage systems, which makes the popularization of object storage difficult.
• Object storage is applicable to scenarios such as big data, IoT, backup and
archive. It provides EB-level capacity and has the highest data durability among
the three types of storage.
• Direct Attached Storage (DAS) connects an external storage device to an
application server through the SCSI or FC interface, making the storage device
part of the server. In this case, the data and operating system are not separated.
• Network Attached Storage (NAS) uses TCP/IP, ATM, and FDDI to connect storage
devices, switches, and clients, and all these components form a private storage
network. NAS integrates storage devices, network interfaces and Ethernet
technology and stores data directly over Ethernet, which separates the storage
function from the file server.
• Storage Area Network (SAN) is a private storage network that connects storage
arrays and servers through switches.
• As data grows exponentially, storage of massive amount of data imposes great
pressure on local storage and brings heavy burden to existing storage systems. To
relieve the pressure, we have to adopt distributed storage and distributed file
systems.

• How can we ensure high performance and high availability of distributed storage?

▫ In addition to the backup, active-active, and multi-active architectures in

the traditional architecture, multiple data copies are stored in the system to
ensure high reliability and availability of the distributed storage system. If a
storage node becomes faulty, the system can automatically switch the
node's service to other nodes, achieving automatic fault tolerance. The
distributed storage system leverages replication protocols to synchronize
data to multiple storage nodes and ensures data consistency between
copies. A piece of data has multiple copies, among which there is only one
primary copy, and the rest are backup copies. Consistency is used to ensure
data integrity when data is replicated from the primary copy to backup
copies.
• Elastic Volume Service (EVS) provides persistent block storage for ECSs and BMSs.
With data redundancy and cache acceleration techniques, EVS offers high
availability, strong durability, and low latency. Users can format an EVS disk,
create a file system on it, and store data persistently.

• SFS is a network attached storage (NAS) service that provides scalable, high-
performance file storage. With SFS, you can enjoy shared file access spanning
ECSs, BMSs, and containers created on CCE and Cloud Container Instance (CCI).

• Object Storage Service (OBS) provides a stable, secure cloud storage that is
scalable, efficient, and easy-to-use. It offers REST APIs and allows users to store
any amount of unstructured data in any format.
• B. The bridge only connects the network port of a VM to a physical network port
to access external networks through a physical server. The NAT translates
network addresses to enable VM traffic to access external networks through a
physical network port.

• C. Currently, there are three mainstream storage types: block storage, file storage,
and object storage. Tapes are also a type of storage media, but they are only
used in some backup and archive scenarios.