H13-531_V2.0 HCIE-Cloud Computing (Written) V2.0 Questions and Answers

Create snapshots to protect snapshot point-in-time data

Rollback snapshot restores the data at the snapshot point in time to the original disk

Delete the snapshot, delete the snapshot completely, cancel the data protection at this point in time

Snapshot sharing, share snapshots to specified tenants

Data synchronization and access across multiple sites

high cost

Stateless access control node to handle massive user requests

Massively linear scaling

FusionSphere OpenStack OM

FusionManager

FusionCompute

FusionStorage

Use the NAT function of the router to open up the network between the source and destination virtual machines.

Configure the routing function on the rainbow to connect the network between the source and destination virtual machines.

Configure nat on rainbow; connect the network between the source and destination virtual machines.

Use the routing function of the Layer 3 switch to open up the network between the source and destination virtual machines.

OneDVSThere can be multiple port groups, aVMCan be protected by multiple security groups.

A virtual machine NIC is connected to a virtual port of the port group, and a virtual machine NIC is associated with a security group.

Security groups are based on port groups to protect virtual machines.

port group isDVSA collection of one or more virtual ports, while a security group is a collection of one or more rules Correct

Change the status of the disk from available to unavailable

Change the mode of the disk from slave to independent-persistent

C Change the capacity of the disk from 80G to 8G

D Change the configuration mode of the disk from thin to normal

deleteVLANsegment, there must be a startVLAN IDand endVLAN ID,Delete the entireVLANpart

if there isVLANThe pool can create a port group.

OneDVSdifferent inVLANbetween poolsVLAN IDcannot overlap

createVLANStart must be set when poolingVLAN IDand endVLAN ID

0Sstate

Reliability inspection

Security Inspection

Availability inspection

Share Mode

Copy Mode

Link Mode

Pool Mode

True

False

10Hour

10minute

10second

0second

physical location topology

VDCtopology

Service Application Topology

VPCtopology

After the storage link is interruptedHA

After the network connection is interruptedHA

After virtual machine blue screenHA

After a host failureHA

Whether multiple hosts belong to the same cluster depends on whether they use the same shared storage.

Hosts can be moved between different clusters.

Virtual machines can be migrated across clusters.D. A cluster is composed of multiple hosts and their associated storage and network resources.

Cloud hard disk instances can be mounted on running cloud host instances

existVDC (Virtual Data Center)Unified management of computing, storage and network resources

In the general cloud hosting service, support advanced properties such as custom hostname, key pair, specified password, etc.

elasticityIPThe address is a static publicIPaddress that can be elasticIPThe address is associated with the cloud host

True

False

The cluster has DPM enabled

The IP address of the business virtual machine is not the same as the IP of the CNA host—the network segment

The memory multiplexing configuration of the source host and the target host is inconsistent

The operating system of the business virtual machine is inconsistent with the operating system type or version of the CNA host

True

False

File method

to share

local way

remote

ITA sends username and password to virtual machine

HDC checks whether licnese is satisfied

The virtual machine uses the domain account password to log in to AD for authentication

TC establishes a login connection with the virtual machine

nova-compute in the nova project

neutron-server in the neutron project

neutron-13-agent in the neutron project

nova-scheduler in the nova project

The cloud disk has been mounted to the cloud host

The cloud hard disk is the system disk

The cloud hard disk has been transferred

Cloud Drive is created for the application

When FusionCloud 6.3 is installed, the deployment scenario needs to be determined.

FusionCloud 6.3 is installed uniformly using FCD.

Fusioncloud 6.3 requires manual installation of all components.

Fusioncloud 6.3 needs to determine the cloud service type during installation, and the installation is completed at one time.

445, the destination virtual machine mounts the LiveCD

7443, create a destination virtual machine

443, send a snapshot creation command

1880, log in to the migration server remotely and issue a migration task

IMCMode configuration

disk hot add

clusterHAconfigure

Computing resource scheduling configuration

The complete replication of desktops is based on virtual machine-level isolation, with high security

When creating a complete copy of a virtual desktop, the system will assign an independent system disk to the virtual desktop

space, and copy the virtual machine template completely to the system disk

The complete replication virtual desktop has a separate system disk and user data disk

Use virtualization technology and remote desktop delivery technology to fully replicate desktops, and use virtual

Virtualization technology to virtualize servers and storage into elastic virtual hosts

Routine backup

Before the server replaces the hard disk

Before upgrading the service instance

Before major business adjustments

True

False

eBackup Workflow calls the Swift interface to create a volume snapshot and activates it with snapshot consistency

eBackup Workflow calls the Cinder backup interface to notify eBackup Server to perform volume backup

eBackup Workflow calls Nova interface to obtain cloud server metadata

CSBS Karbor initiates task scheduling and notifies eBackup Manager to perform full backup of the cloud server

TRUE

FALSE

able to passsshLog in to the user VM to collect logs.

can be collectedGuessDBThe database service check log and database operation log.

able to passgandalfusersshLog inLinuxThe system infrastructure virtual machine virtual machine, and then copy the component logs to the local.

LinuxSystem logs are generally in /var/logUnder contents.

True

False

VLB

ITA

HDC

WI

nova-api

KeyStone

HAProxy

RabbitMQ

The target host has failed, has been rebooted, or has entered maintenance mode

source host and target hostCPUThe amount is inconsistent.

The source host and the target host have network interruptions or network failures

The virtual machine is bound to the source host

Unable to meet requirements

Virtual Machine Live Migration Changes Datastore

Virtual machine live migration changes hosts and datastores

Virtual Machine Live Migration Change Host

TRUE

FALSE

Management side operation log

Operation log

Security log

Tenant operation log

Gateway basic information

Virtual Machine Usage Information

Basic information of components

User connection information

Memory over-allocation

DRS

NUMA Affinity Scheduling

CPU over-allocation

GPUdesktop

Linked-clone desktop

full memory desktop

full copy desktop

Glance

Neutron

Nova

Ironic

True

False

The cost of traditional storage devices increases significantly with performance improvement.FusionStorageSupports converged deployment, saves costs, recycles old equipment, and protects investment.

To expand the memory of traditional storage devices, increase the number of controllers and replace devices, whileFusionStorageCapacity increases linearly with

performance

CTraditional storage device hard disk failure does not require management, automatic data reconstruction and recovery,FusionStorageHard disk failure needs to be replaced immediately, real-time manual recovery

Traditional storage device data reconstruction is slow,FusionStorageData is distributed to servers and cabinets, and data reconstruction in the full resource

pool is faster after a fault.

Recommended network card typevirtio'

When selecting a storage resource, you can choose to passSSDCard,UMapped data storage resources such as disks.

If the virtual machine is deployed in active and standby mode, it is recommended to deploy the two virtual machines in different parts of the management cluster.CNAson the node.

When choosing storage resources, it is recommended to use high-performance storage instead ofNL-SASorSATA etc. Low performance

OneVPC there is only one subnet below

VPCAllows computing resources to be isolated over the network

A user can have multipleVPC

VPCis to logically isolate the network

If you want to perform some major operations during the troubleshooting process, you should back up the relevant databases, alarm information and log files, etC.

Fault handling should comprehensively consider the possibility and time cost of the corresponding operation to restore business success.

The principle is to restore the system as soon as possible.

When determining the fault handling party, you should first evaluate the impact and give priority to ensuring the normal operation of the business.

VDCadministrator andVDCThe salesperson can define the approval process.

The approval process supports real-time modification of wells to take effect in the release state.

Each level of approval process can specify one or more approvers, as long as any approver approves, the approval of this level is passed.

The approval process can set up multi-level approval, up to5class correct.

There is no virtual machine in the security group. If there is a virtual machine in the security group, you need to move the virtual machine out of the security group

The authorization behavior between the security group to be deleted and other security groups does not affect the removal operation

Except for a security group to which the virtual machine belongs, other security groups are not affected by the virtual machine

After the security group is deleted, all security group rules in the group are deleted at the same time

TRUE

FALSE

support pairFusioncomputevirtual machine backup

Accurate backup of virtual machine desktop files

Incremental backup to disk

You can do a full backup of the virtual machine and a permanent incremental backup

Lease higher bandwidth for initial data synchronization

WillBCManager eReplicationThe host-based remote replication disaster recovery scheme is changed to storage remote replication disaster recovery

Add links and configure load balancing between the active and standby sites

WillBCManaaer eReolicationBased on the storage remote replication disaster recovery scheme, the

50piece2TB SATADisks form storage pools. internal useRAID5type

100piece1TB SATADisks form a storage pool for internal useRAID5type

50piece2TB SATAAdministrative disks form a storage pool for internal useRAID6type

100piece1T8 SATADisks form storage pools. internal useRAID6Beautiful

True

False

Take the host out of maintenance mode

Recreate a virtual machine with the same specifications

Check the remaining storage space to ensure that there is enough storage space

Repair the host network link

wrong

right

True

False

True

False

Adress ticket

Login ticket

Network ticket

Token

Multiple hard disk failures in a storage pool in a short period of time

The hard disk where the last copy of the three copies of data resides orCacheUnreadable failure occurred

Storage network exception

Not enough hard drive space

Backup Workflow

backupBackup agent

arbor

Backup Manager

TRUE

FALSE

ZK, which one starts first and which is the main ZK

The MDC selected by more than half of the ZKs is the main MDC

MDC and zK are not one-to-one relationship, the master MDC and master ZK must be on the same server

ZK is deployed on the control node

Rainbow Migrator Linuxmust be deployed

Decide whether to deploy or not based on the source virtual machine to be migratedRainbow Migrator

Rainbow Servermust be deployed

Rainbow Migrator WindowsThe correct answer must be deployed

Perform operations such as desktop group management, user association with virtual desktops, etc. according to the request sent by the ITA.

When the user accesses, interact with the WI, provide access information, and support the completion of the entire user access process.

interacts with the HDA in the VM, and collects the virtual machine status and access status reported by the HDA.

A single set of HDC desktop controllers can manage 1000 virtual desktops.

Disconnect the desktop first, and re-login to the desktop through WI

Contact the administrator to shut down and restart the CNA host where the virtual machine is located

Uninstall the incompatible input method and reinstall the compatible input method

Check whether the firewall is closed

RPOis the recovery time objective

RTOis the recovery point objective

RTOis the recovery time objective

RPOis the recovery point objective

Rainbow Server must be deployed.

Rainbow Migrator Linux must be deployed.

According to the source virtual machine to be migrated, decide whether to deploy RainbowMigrator.

Rainbow Migrator Windows must be deployed.

When backing up DHCP, you need to configure DHCP information in the "Alarm Component" of FusionAccess

When backing up AD, in FusionAccess "System Management > Initial Configuration > Domain/OU", the

"Enable Backup" is set to "Yes"

When backing up DNS, you need to configure DNS information in FusionAccess "Alarm Component"

When backing up VLB components, you need to configure VLB information in FusionAccess "Monitoring Components"

Check the log for troubleshooting, and retry the entire task after completion

View the specific error log. If manual configuration is required, skip this step after the configuration is complete.

Check the logs for troubleshooting, then skip this step

skip directly

The memory is recommended to be greater than or equal to the source specification.

Configure at least one network card and use the system to assign an IP address, which cannot be modified manually.

The CPU is recommended to be greater than or equal to the source specification.

The destination virtual machine system disk needs to be configured in MBR format.

subnet

network

port

elasticityIP

The same virtualization is divided into several Azs using the principle of maximum computing pool and storage pool, that is, if there are 2 computing pools

All pools are connected to the same set of storage pools and are divided into 2 Azs: If there is 1 compute pool connected to 3 sets of storage pools,

Divide into 3 Az

Openstack can be divided into multiple different bare metal resource pools (AZ) C. The division of host groups is ultimately reflected in the choice of specifications when creating cloud hosts. Specification of labels required and host

If the tag of the group matches, selecting a certain specification is equivalent to selecting the host group with the corresponding tag.

AZ is included in Region, Az cannot cross Region

True

False

TRUE

FLASE

Backup protocol supports local, via FTP/SFTP to third-party server and OBS

In order to ensure the stable operation of the system, backup and recovery of management data can ensure that the impact on the business is reduced to

lowest

For the business recovery of CloudOpera, the same business can choose multiple backups to restore the actual file information

The system supports both automatic backup and manual backup of data

DRS

memory hot add

snapshot

Thermal migration

To create a virtual machine instance fromGlanceQuery image information

Image files do not need to be changed frequently, and are more suitable for storing inSwiftmiddle.

Glancemust useSwiftAs a backend storage for images.

The volume data of the virtual machine instance is updated frequently and is more suitable for storing inCinder

DNSThe configuration information of the forward lookup zone in the server is modified

twoHDCTime out of sync between servers

virtual machine withHDCThe network status is abnormal.

DHCPComponent exception

On the same physical host, allGuest Physical Memorycan be larger than the physical host'sHost Physical Memory

On the same physical host, allGuest Physical MemoryMust be less than or equal to the physical host'sHost Physical Memory

In a virtual machine, all applications correspond toGuest virtual memorycan be larger than the virtual machine'sGuest Physical Memory

Already in a virtual machine, all applications correspond toGuest Virtual MemoryMust be less than or equal to the virtual machine's Guest Physical Memory

existFusionAccess PortalThe interface finds the virtual machineID ,login toHDCOn the component background, enter the corresponding

command to migrate.

existFusionAccess PortalOn the interface, find the virtual machine to be migrated, and click the "Migrate" button.

existFusionCompute Portalinterface, find the virtual machine to be migrated, and click the Migrate button in the Virtual Machines taB.

existFusionAccess PortalThe interface finds the virtual machineID ,login toWIOn the component background, enter the corresponding command line to migrate

storage deviceLicensenot loaded

CHAPAuthentication failed

storageIPand port filling error

IP-SANThe mapping relationship is not configured on the device

4

3

2

5

Applications requesting virtual memory can be0Shenzhen begins.

Memory virtualization technology can achieve memory over-allocation.

FusionComputeSupports memory hardware-assisted virtualization technology to reduce memory virtualization overheaD.

The virtual memory requested by the application is contiguous.

Local high availability can use real-time mirroring and synchronous replication solutions. Due to the close bandwidth and distance, it usually requires RPO=0.

the remote disaster recovery party can choose the remote active-active and remote active-standby solutions according to different needs

For intra-city disaster recovery, you can choose intra-city active-active and intra-city active/standby solutions according to different needs.

Local high availability deploys services in the same computer room in close proximity.

Check all hosts when initializing storage resources

When the shared storage type is IP SAN, ensure that the storage network of the host and the business interface of the storage are connected. C. Check the corresponding host when adding data storage

Check all hosts when associating storage devices

Physical host resource utilizationTOPNUsed to count the resource usage trends of all physical hosts in the system.

Cloud platform management system resource utilization analysis is used to count the computing resource utilization of each cloud platform management system.

The computing resource utilization trend is used to collect statistics on the average usage of computing resources in a certain period of time in regions, AZs, and clusters.

the aggregation analysis of elastic cloud server resource utilization is used to count the resource usage of elastic cloud server in area, availability zone, and cluster dimension

When using snapshot comparison technology for incremental backup, it is necessary to retain snapshot data on the production side

Use CBT technology for incremental backup, which saves storage space compared to snapshot comparison.

CBT technology will not be used during the first full backup

When using CBT technology for incremental backup, snapshot technology will not be involved

The network cable connected to the host network port is loose

The VRM clock synchronization process is abnormal

The network cable of the switch connected to the host is loose or damaged

The host is restarting

Only virtual machines whose allocation status is "allocated" can restore the system disk

The FusionAccess system supports multiple virtual machines of the same virtual machine type to restore the system disk at the same time

After restoring the system disk, the linked clone virtual machine only retains the data of the data disk, and the data of the system disk will be lost

Only the virtual machine whose running status is "running" can restore the system secret

FusionSphere OpenStack

Vmware Center Server

FusionCompute

VMware ESX/ESXi Server

TRUE

FALSE

Migrate according to the order of migration risk from high to low

Migrate in order of migration complexity from easy to difficult

Migrate according to the order of the migration allowable interruption time from short to long

Migrate according to the order of the impact of the migration on the business from small to large

HBA card is required when connecting to FC--SAN. In this scenario, cinder-volume can be deployed in virtual In the virtual host

There is no need to deploy the image-cache partition when deploying the cinder-volume host

C A set of cinder-volume (at least one node) connected to only one storage pool

cinder-volume can only be deployed on the control node

If the amount of file data in the database is large and the downtime is short, you can re-issue a file at the target end.

After the virtual machine is installed, reinstall the Oracle database, and use the RMAN backup and recovery method for data migration

If the database uses a file system to store data files, and the amount of data is not large, the downtime window is sufficient,

You can choose to migrate the operating system and data by migrating the entire virtual machine at one time after the business is stopped.

database

If the amount of file data in the database is large, you can choose to stop the business and use Rainbow to complete the full amount at one time.

The way to migrate virtual machines to migrate operating systems and databases

If the amount of file data in the database is large and the downtime is long, you can re-issue a file at the target end.

After the virtual machine is installed, reinstall the Oracle database and use Rainbow for data migration

Security group is a virtual machine port-level security service of VPC, supports black and white lists, and can be used to control Control traffic to and from one or more virtual machines

is a subnet-level and VPC-level security service of the VPC

By setting up a virtual firewall, adding an additional security layer to the VPC can be more flexible and hierarchical

Safety management.

Support black and white list, which can be used to control the traffic in and out of one or more subnets

OBS

EIP

VAPP

AS

The host where the virtual machine is located has completed the huge page configuration and takes effect

The cluster where the virtual machine is located has been set to a high-performance cluster

Turn on the IMC mode of the cluster where the virtual machine is located

A distributed switch of the user-mode switching mode type has been created

Stacking mode

Active and standby mode

LACP horizontal based on source and destination MAC

Polling-based load sharing

Data Backup > Migration Tool Installation > Data Synchronization > Data Verification Test > System Migration > System Verification

Test > Difference Data Sync

Data Backup > Migration Tool Installation > System Migration > System Verification Test > Data Synchronization > Data Verification

Test > Difference Data Sync

Data backup > Migration tool installation > Data synchronization > Data verification test > System migration > System verification

Test > Differential Data Same > Data Backup

Data Backup > Migration Tool Installation > System Migration > System Verification Test > Data Synchronization > Data Verification

Test > Differential Data Synchronization > Data Backup

Compared with the complete copy template, using the fast encapsulation template to issue virtual machines is faster and more efficient

There is no essential difference between full reproduction and rapid distribution

using sysprep package

Rename and add domain by the application in the virtual machine

Configuring the cluster HA strategy is to automatically adjust the resource load balancing on different hosts

Dedicated failover hosts refer to: setting a specified number of hosts in the cluster to fail, the system

Periodically check that there are enough resources left in the cluster to fail over the virtual machines on these hosts

To allow host failure in the cluster refers to: Reserving a designated host as a dedicated failover host, when the host is used as a failover host, ordinary virtual machines are prohibited from starting, migrating, waking up, and snapshot recovery on the host

HA resource reservation is to reserve CPU and memory resources according to the configured values ​​in the entire cluster

The user does not configure the NTP clock source, the user can check whether the

If there is any related alarm, follow the alarm handling help to configure the NTP clock source and clear the alarm

This phenomenon is caused by the inconsistency between the time of the NTP clock source and the local PC. The user can modify the time of the local PC.

time, consistent with the NTP clock source

When the key is updated, the business process on the node will be restarted, which will cause abnormal monitoring of the virtual machine.

The phenomenon is caused by the abnormal monitoring process. The user can use "PuTTY" to log in to the main machine as the gandalf user

VRM, restart monitoring process

Migrating Alibaba Cloud's ECS to HUAWEI CLOUD

Migrate the running virtual machine to FusionCompute.

Migrate the virtual machine running in the Beijing data center to the virtual machine running in the Shanghai data center through the Internet on FusionCompute.

The Rainbow server can access the source and destination virtual machines, but cannot access the management network of FusionCompute network.

Cancel the automatic mount function of the file system and remount the file system immediately

Stop the business operation on the faulty storage device

Repair read-only volumes via fsck

mkfs. ext4-fn device command” to check the faulty device information