• Flat network: A flat network is connected to the network where the host is located and has direct access to the Internet. VM instances in a flat network can access public networks by using elastic IP addresses.
• VPC network: A VPC network is a private network where VM instances can be created. A VM instance in a VPC network can access the Internet through a VPC vRouter.

• The VM console allows you to conveniently monitor and manage VM instances. Note that you must have the corresponding permissions before you can log in to the VM console.
• A VM instance supports IP/MAC/ARP anti-spoofing. Abnormal protocol access sent by the VM instance at the data link layer of a host is isolated to block MAC/ARP spoofing and achieve IP anti-spoofing at the network layer.
• You can create an image for a VM instance. The image contains all of the information about the VM instance. You can quickly copy the corresponding resources through the image.
• You can create a snapshot for a VM instance. Before you perform a business-sensitive operation on the VM instance, you can schedule snapshot creation at specified time points to record the state of the VM data. This allows rollback in case of breakdowns.
• A VM instance supports encrypted storage of plain text passwords to protect sensitive data on the VM instance.
• You can set a delete policy for a VM instance, including Direct, Delay, and Never. When you delete a VM instance in the UI, a dialog box is displayed to remind you of the consequences of the deletion. You must acknowledge the risks before you can delete it. This helps to reduce the risks caused by misoperations.
• A VM instance supports role-based access control and permission management.
• A VM instance supports operation logs and auditing, which can meet the needs of security analysis, intrusion detection, resource change tracking, and compliance auditing.
• The Cloud is able to securely segregate the individual guest VM instance.

• A VM instance allows you to modify its CPU and memory online and expand the attached root volumes and data volumes online. You can modify the VM configurations as needed.
• A VM instance supports auto-scaling. The Cloud can automatically trigger VM auto-scaling or self-healing according to business changes.
• A VM instance supports multiprocessing for virtual processors.

VM instances integrate the business of an enterprise on the cloud and migrate the information service system from traditional physical servers to VM instances. This helps to improve the resource utilization and reduce repeated investments. VM instances realize the rationalized scheduling of resources through intelligent load balancing services. In addition, the VM HA feature can deal with various exceptions to ensure business continuity of VM instances.

For modern IT enterprises, the deployment and approval of a development and testing environment is time-consuming, which severely lengthens the business launch cycle. With VM instances, resources can be allocated online, which helps to establish or recover a development and testing environment in seconds. This accelerates the business launch. In the same resource pool, an enterprise can use the environment encapsulated in a VM instance for development at day time and for automated testing at night. After an application development is completed, resources occupied by the corresponding VM instance can be quickly released and assigned to other projects. You can plan resource configurations in advance, allowing projects to apply for needed resources which are assigned to them in time.

For enterprises that cannot migrate PaaS or SaaS services to the public cloud, they can use ZStack Cloud to build a private cloud environment and deploy the PaaS or SaaS services on VM instances. The flexibility, stability, and high concurrency characteristics of the VM instances help to ensure the security, stability, and high-efficiency of the enterprises.

In recent years, network attack defense is tilted towards attackers. Enterprises are facing severe cloud security challenges as various high-risk vulnerabilities, APT targeted attacks, and computer viruses emerged in an endless stream. By using VM instances, enterprises can build a completely isolated security rehearsal environment and ensure the business security through monitoring and alarming, log auditing, vulnerability management, anti-virus and other means.

The name must be 1 to 128 characters in length and can contain Chinese characters, letters, digits, spaces, hyphens (-), underscores (_), periods (.), parenthesis (), colons (:), and plus signs (+) and cannot begin or end with spaces.

• Basic Offering: Set the VM offering by selecting an existing instance offering and disk offering.
  • Instance Offering: Select an existing instance offering.
    • Reserve Memory: The minimum memory that ensures the normal running of the VM instance. By default, this feature is disabled. If you enable this feature, make sure that the reserved memory is less than the memory defined in the instance offering.
  • Image: Select an existing image.
    Note: If you select an ISO image encapsulating a Windows operating system with Virtio enabled. The VM system will automatically load Virtio drivers. You need to manually in install the drivers when install the operating system. After you reboot the VM instance on UI, the drivers are automatically detached.
  • BIOS Mode: Select a BIOS mode for the VM instance. This parameter is required if your image is in the ISO format. By default, the VM BIOS mode is consistent with the image. You can also manually select another one.
  • Root Disk Offering: Select an existing disk offering for the root volume of the VM instance. The root disk offering defines the root volume information such as the size and disk bandwidth.
    • If you select a raw or qcow2 image, this parameter is optional. If you do not select a disk offering, the size of the root disk offering is consistent with the size of the selected image.
    • If you select an ISO image, this parameter is required.
  • Data Volume: Optional. Choose whether to create data volumes and attach the volumes to the VM instance. You can attach a default maximum of 24 data volumes to a VM instance. The maximum number can be changed by modifying the global setting Maximum Data Volume. By default, no data volumes is created and attached to a VM instance.
    If you choose to attach a data volume, click Attach Data Volume and set the following parameters:

    • Data Disk Offering: Select an existing disk offering for the data volume of the VM instance. The data disk offering defines the data volume information such as the volume size and disk bandwidth.
    • Quantity: Enter the number of data volumes created from the selected data disk offering that you want to attach to the VM instance.
    • Enable VirtIOSCSI: Optional. Choose whether to use VirtioSCSI bus to create a SCSI data volume.
      Note:

      • By default, VirtioSCSI is enabled if a shared storage is used, indicating that you can create VirtIO SCSI volumes.
      • If a LocalStorage primary storage is used, you can enable VirtioSCSI on the volume details page.
      • VirtIO SCSI volumes support multiple I/O queues, which can be identified as IDs (WWN).
• Custom Offering: Set the VM offering by customizing the following VM configurations:
  • CPU: Set the number of CPU cores of the VM instance. Valid values: 1 to 1024, integer.
  • Memory: Set the memory size of the VM instance. Valid values: 16 MB to 1000 TB, integer. Unit: MB, GB, and TB.
  • Reserve Memory: The minimum memory that ensures the normal running of the VM instance. By default, this feature is disabled. If you enable this feature, make sure that the reserved memory is less than the memory defined in the instance offering.
  • Host Allocation Strategy: The strategy used to allocate a host to a VM instance. This strategy is consistent with the global setting Host Allocation Strategy . You can modify this global setting to specify a host allocation policy for the VM instance.
  • Image: Select an existing image.
  • Root Volume: Set the size of the root volume. Valid values: 1 MB to 1024 TB, integer. Unit: MB, GB, and TB.
    • If you select a raw or qcow2 image, this parameter is optional. If you do not select a disk offering, the size of the root disk offering is consistent with the size of the selected image. If you customize a size, make sure that the size is no smaller than the selected image size.
    • If you select an ISO image, this parameter is required. Make sure that the size is no smaller than the selected image size.
  • Data Volume: Optional. Choose whether to create data volumes and attach the volumes to the VM instance. You can attach a default maximum of 24 data volumes to a VM instance. The maximum number can be changed by modifying the global setting Maximum Data Volume. By default, no data volumes is created and attached to a VM instance.
    If you choose to attach a data volume, click Attach Data Volume and set the following parameters:

    • Data Disk Offering: Select an existing disk offering for the data volume of the VM instance. The data disk offering defines the data volume information such as the volume size and disk bandwidth.
    • Quantity: Enter the number of data volumes created from the selected data disk offering that you want to attach to the VM instance.
    • Enable VirtIOSCSI: Optional. Choose whether to use VirtioSCSI bus to create a SCSI data volume.
      Note:

      • By default, VirtioSCSI is enabled if a shared storage is used, indicating that you can create VirtIO SCSI volumes.
      • If a LocalStorage primary storage is used, you can enable VirtioSCSI on the volume details page.
      • VirtIO SCSI volumes support multiple I/O queues, which can be identified as IDs (WWN).

• HA Mode: Optional. Specifies whether to enable auto restart if VM instances are unexpectedly stopped or are errored because of errors occurred to compute, network, or storage resources associated with the VM instances. Valid values: None and NeverStop.
  • None: The VM instance does not reboot automatically after an unexpected stopping or error.
  • Neverstop:
    • The VM instance attempts to reboot automatically after it is unexpectedly stopped because of its own errors.
    • The VM instance attempts to migrate to another host automatically if errors occur to the compute, network, or storage resource it is residing on.
    • The VM instance does not reboot automatically if you stop it manually, including:
      • Manually perform the stopping VM instance, force stopping VM instance, and powering off VM instance actions on the UI.
      • Manually run the shutdown, poweroff, and halt commands in the VM OS.
      • Create a scheduled job to trigger the VM shutdown as planned.
  Note:

  • By default, a newly created VM uses VM HA Mode Default value in the platform setting. You can modify this value for a VM instance individually during or after the creation. After the modification, the VM instance is not affected by VM HA Mode Default Value in the platform setting.
  • If you choose NeverStop, make sure that you have enabled HA Policy in Platform Setting. Otherwise, the NeverStop mode does not take effect.
• Sync with Host BIOS Time: Optional. Choose whether to sync Windows VM time with the host BIOS time. If enabled, Windows VM time is the same as that of the host. By default, the sync is enabled.
  Note:

  • If you enable time synchronization for a Windows VM instance, the VM time is synchronized with the host BIOS time automatically.
  • If you disable time synchronization for a Windows VM instance, the VM time is not synchronized with the host BIOS time.
  • If you modify the time synchronization setting for a Windows VM instance, the setting takes effect after the VM instance reboots.
  • The time synchronization setting does not take effect on Linux VM instances.
  • You can enable time sync in Global Setting for all Windows VM instances on the Cloud. To enable time sync in Global Setting, set Sync with Host BIOS Time to Sync. By default, the time sync global setting is enabled.
  • If you specifically set time sync for a Windows VM instance, the global setting does not take effect on the VM instance.
• Resource Priority: Optional. Set a high resource priority for the VM instance so that VM instance can compete for more resources in case of resource contention.
  Note:

  • When resource contention occurs due to high host workloads, priority is given to allocating resources to VM instances with the high resource priority.
  • We recommend that you set a high resource priority for vital VM instances.
• VM Scheduling Group: Optional. Add the VM instance to a VM scheduling group.
  • A VM instance can be added to only one VM scheduling group. After the addition, the VM instance will be scheduled based on the scheduling policy associated with group.
  • The scheduling policies associated with a VM scheduling group can be classified into the following four types: VM Exclusive from Each Other, VM Affinitive to Each Other, VMs Affinitive to Hosts, and VMs Exclusive from Hosts.
• USB Redirection: Optional. Set USB redirection for the VM instance.
  Note:

  • The Cloud currently supports redirection of multiple USB devices. If you want to use the VDI feature, you can redirect the USB device on the VDI client to the VDI VM instance.
  • The setting takes effect after you restart the VM instance.
• Anti-Spoofing Mode: Optional. Enable the anti-spoofing mode for the VM instance.
  Note:

  • This mode provides IP/MAC anti-spoofing and ARP anti-spoofing for the VM instance. If enabled, the VM instance can only communicate with outside networks by using the IP or MAC addresses allocated by the Cloud.
  • You can set whether to enable the anti-spoofing mode globally by using the following method:

    Go to Settings > Platform Setting > Global Setting > Basic > Resource Center > Resource Pool > VM Instance, locate Anti-Spoofing Mode, and enable or disable the switch as needed. Default: false.

  • By default, the anti-spoofing mode of a single VM instance is strictly controlled by that in the global setting.
  • If anti-spoofing is disabled in the global setting, you can enable it for a single VM instance on the VM details page.
  • Anti-spoofing mode is not compatible with IPv6 addresses. We recommend that you do not enable it for the VM instance if you decide to configure an IPv6 address for it because this will make the VM instance fail to communicate through the IPv6 address.
• Instance Offering Online Modification: If enabled, you can modify the number of vCPUs and memory size of a VM instance online
  • This setting is consistent with the global setting Instance Offering Online Modification by default. However, if you specifically turn on/off this switch for this VM instance, the global setting does not take effect on this VM instance.
  • We recommend that you do not modify the number of vCPUs and memory size of a Windows-based VM instance in the production environment.
• XML Hook: Attach an XML Hook to insert or modify parameters in XML files of the VM instance, thus realizing custom configurations and specialized functionalities.
  Note:

  • The XML Hook takes effect automatically after the VM instance is created and started.
  • If you detach the XML Hook from the VM instance in future, the configurations that are added or modified by the XML Hook will revert to original status.

• Network: Select an L3 network for the VM instance. Supported network types: VPC network, public network, and flat network.
  Note:

  • If the L3 network is enabled with the DHCP service, the NIC acquires an IP address in a DHCP mode. The network configurations are deployed by the DHCP service and take effect directly.
  • If the L3 network is disabled with the DHCP service, the NIC acquires an IP address in a Static mode. You need to install GuestTools for the VM instance after the creation and sync the configurations to the NIC to take effect. If the VM image has encapsulated GuestTools, the network configurations take effect directly after the VM instance is created.
• Make Default: Choose whether to set the network as the default network of the VM instance.
• Enable SR-IOV: Optional. Choose whether to use SR-IOV to generate a VF NIC and pass it through to the VM instance. By default, SR-IOV is disabled. You can enable it if you have hardware resources that satisfy related requirements.
  Note: To enable SR-IOV, note that:

  • Make sure that SR-IOV network acceleration mode is used for the L2 network from which the L3 network is created. Otherwise, SR-IOV cannot be enabled.
  • Enabling SR-IOV generates an NIC of the VF type and passes it through to the VM instance.
  • After SR-IOV is enabled, make sure that the physical NIC corresponded by the L3 network has an available VF NIC. Otherwise, the VM instance might fail to be created.
  • If you do not enable SR-IOV:
    • If the L2 network from which the VM L3 network is created uses Normal or Smart NIC network acceleration mode, the VM instance is assigned an NIC of the vNIC type.
    • If the physical NIC in the L2 network from which the L3 network is created is a Smart NIC, and the L2 network uses Smart NIC network acceleration mode, the VM instance is assigned an NIC of the vDPA type.
• Assign IP: Optional. Choose whether to manually specify an IP address to the VM NIC.
  • If the L3 network is enabled with the DHCP service, the system can automatically allocate an IP address to the VM NIC according to the IP allocation policy of the L3 network. You can either manually specify an IP address or not. If you choose to manually assign one, set the following parameters:
    • Assign IPv4/IPv6: Specify an IP address on the L3 network for the VM NIC.
      Note:

      • The IP address must be within the network range of the selected network and not be occupied by other resources.
      • By default, the drop-down menu displays 5 recommended IP addresses. If the current network has less than 5 IP addresses, all available IP addresses will be displayed.
      • If you create VM instances in bulk and specify an IP address here, the Cloud will automatically assign a range of IP addresses to these VM instances with the IP address you specify as the start IP. If an IP address is occupied or if IP addresses are insufficient within the continuously assigned IP range, the corresponding VM instance will fail to be created.
    • If the L3 network is disabled with the DHCP service, the system does not automatically allocate an IP address. To configure a NIC IP address, you can either assign one on the Cloud or configure one in the VM instance.
      If you assign an IPv4 on the Cloud, set the following parameters:

      • Assign IPv4: Specify an IPv4 address for the VM NIC.
        Note:

        • Make sure that the IPv4 address has not been occupied on the Cloud.
        • If the L3 network is enabled with IP Address Management, make sure that the IPv4 address is within the network range of the L3 network.
        • If you create VM instances in bulk, the IPv4 address you enter is identified as the start IP and assigned to the first VM instance. The IPv4 addresses following it are continuously assigned to the rest VM instances. When an IPv4 has been occupied, the corresponding VM instance fails to be created.
        • After the creation, install VM GuestTools and sync NIC configurations to make the IPv4 take effect. If the VM image has encapsulated GuestTools, the IPv4 takes effect directly after the creation.
      • Netmask: Set the IPv4 netmask.
      • IPv4 Gateway: Optional. Set the IPv4 gateway.
      If you assign an IPv6 address on the Cloud, set the following parameters:

      • Assign IPv6: Specify an IPv6 address for the VM NIC.
        Note:

        • Make sure that the IPv6 address has not been occupied on the Cloud.
        • If the L3 network is enabled with IP Address Management, make sure that the IPv6 address is with in the network range of the L3 network.
        • If you create VM instances in bulk, the IPv6 address you enter is identified as the start IP and assigned to the first VM instance. The IPv6 addresses following it are continuously assigned to the rest VM instances. When an IPv4 has been occupied, the corresponding VM instance fails to be created.
        • After the creation, install VM GuestTools and sync NIC configurations to make the IPv6 take effect. If the VM image has encapsulated GuestTools, the IPv6 takes effect directly after the creation.
      • Prefix Length: Set the IPv6 prefix length.
      • IPv6 Gateway: Optional. Set the IPv6 gateway.
      If you configure an IP address in the VM instance, the IP address can be read to the Cloud through GuestTools and then managed by the Cloud.

      Note:

      • If the IP address is conflict with IP addresses of other resources on the Cloud, this IP address will not be read and displayed. In addition, an alarm will be triggered.
      • If the L3 network is enabled with IP Address Management, make sure that the IP address you configure is within the network of the L3 network.
      • The IP address assigned on the Cloud is prior to the IP address configured in the VM instance. If you have assigned an IP address to the NIC on the Cloud, the IP address of the same type you configure in the VM instance will not be displayed on the Cloud be default. For example, if you have assigned an IPv6 to the NIC on the Cloud, the IPv6 you configure for it in the VM instance will not be read and displayed on the Cloud. If you want to make this IP displayed can overwrite that one you assign on the Cloud, contact the official technical support to enable the global setting enable.vm.internal.ip.overwrite. This global setting is disabled by default.
• MAC Address: Optional. Choose whether to customize a MAC address for the VM instance. By default, this option is not selected and the Cloud automatically assigns a MAC address to the VM instance. If you choose to configure a MAC address, set the following parameters:
  • MAC Address: Customize a MAC address for the VM instance.
  • Note:

    • If you use the custom configuration, the Cloud assigns a MAC address to the VM instance according to your configurations. In this case, pay attention to avoiding MAC address conflicts.
    • If you create VM instances in bulk and choose to assign a MAC address here, the Cloud automatically assigns MAC addresses to these VM instances with the first assigned MAC address as the start address. If a MAC address is occupied within the continuously assigned MAC range, the corresponding VM instance will fail to be created.
• Security Group: Optional. Associate an existing security group with the VM instance. If the L3 network is disabled with IP Address Management, assign an IP on the Cloud before you can attach security groups.
• EIP: Optional. Associate an existing elastic IP address (EIP) with the VM instance. If the L3 network is disabled with IP Address Management, you cannot attach an EIP.

• System allocation: The Cloud allocates a primary storage according to the preconfigured policy.
• Manual allocation: Select a primary storage manually.

• SE (Security Element) is an embedded security IP core within the processor. It provides foundational software and application systems with secure, compliant, and standardized cryptographic support.
• Enabling this feature allows an SE device from the host to be passed through to the VM instance. Currently, passthrough is supported for SE devices on hosts with LoongArch or Hygon processors.
  • LoongArch: If enabled, you can manually attach an SE device to the VM instance after the creation via the VM's details page -Configuration Info -SE Device list.
  • Hygon: If enabled, the system automatically assigns an SE device to the VM instance when it starts. If you are creating VM instances in bulk, insufficient SE devices cause the corresponding VM creation to fail.
• To enable this feature, make sure that there are available SE devices on the host. You need ensure that the host has an SE device and virtualize it in advance.

• GPU Type: Select a GPU type. Supported: pGPU and vGPU.
  • pGPU: Passthrough a physical GPU device to the VM instance.
  • vGPU: Attach a vGPU device which is virtually divided from a pGPU to the VM instance.
• GPU Attachment Policy: Select a GPU attachment policy.
  • Specify GPU Specification: Select a GPU specification, and the system allocated GPU device(s) to the VM instance according to this specification. If you select this policy, you need set the GPU Count to attach and Detachment Mechanism.
    • Auto Detach: This is the default detachment mechanism. When the VM instance is stopped, the GPU device(s) are automatically detached. When the VM starts, the system re-allocates GPU device(s) to the GPU specification.
    • Do Not Auto Detach: When the VM instance is stopped, it keeps the GPU device(s) attached. However, if the VM is stopped unexpectedly, it may fail to automatically start even though its HA mode is NeverStop.
  • Specify GPU Device: Select a GPU device and attach it to the VM instance directly.
• Load GPU Specification: Allocate a CPU device according to the selected specification when you create a VM instance.
• Auto Detach GPU Device: Choose whether to uninstall a GPU device automatically after the VM instance is stopped.
  • If a pGPU specification is used, this option is not selected by default. If a vGPU specification is used, this option is selected by default.
  • If selected, the GPU device will be automatically uninstalled after the VM instance is stopped. When the VM instance is restarted, a new GPU device will be allocated to the VM instance according to the GPU specification.
  • If not selected, the GPU device will be retained after the VM instance is stopped. When the VM instance is restarted, the existing GPU device will be used.
• Load GPU Device: Attach the GPU device you selected when you create a VM instance.

Note:

• You can attach multiple pGPU devices or only one vGPU device to a VM instance at a time.
• You cannot attach both pGPU and vGPU devices to a VM instance at the same time.
• You can attach GPU devices to a VM instance from the host where the VM instance is located. Currently, you cannot attach GPU devices to a VM instance across hosts.

• Not Set: The vCPUs of the VM instance are not pinned to specific pCPUs but are assigned to pCPUs by the operating system.
• By NUMA Topology: Pins vCPUs to pCPUs based on the host pNUMA topology. You can implement Manual Pinning or Smart Pinning.
  • Manual Pinning allows you to customize vCPU-pCPU binding for all vCPUs.
  • Smart Pinning assigns a vCPU of a VM instance to a pCPU on a host pNUMA node by descending order of the pNUMA node ID. If all pCPUs on the pNUMA node are pinned, a pCPU on the next pNUMA node is to be pinned. If all pCPUs of the host are pinned but some vCPUs are not assigned to pCPUs, each vCPU is assigned to a pCPU starting again from the pNUMA node with the maximum node ID.
  • You can pin a vCPU to one or more pCPU or pin one or more vCPU to a pCPU.
  • The utilization of each pCPU in the past 15 minutes is displayed. You can pin a vCPU to the optimal pCPU based on the utilization.
  • Before you can select By NUMA Topology, you must select a host for the VM instance.
• By Entry:
  • You can click Add CPU Pinning to specify more CPU pinning rules.
  • Enter the vCPU range in the left text box and the pCPU range in the right. Note that the - symbol indicates the value range, while the ^ symbol indicates that a value is not included. If you specify multiple entries in a rule, separate each entry by using a comma (,). The following are some examples:
    • 0-2 indicates CPU 0, CPU 1, and CPU 2.
    • ^2 indicates that CPU 2 is not included.
    • 0-2,^2 indicates CPU 0 and CPU 1.
    • 1-7,^2,^3,^4,10 indicates CPU 1, CPU 5, CPU 6, CPU 7, and CPU 10.

Note:

• The vCPU range depends on the selected instance offering of the VM instance, while the pCPU range depends on the selected cluster or the number of pCPUs of the selected host.
• If a vCPU has multiple CPU pinning rules, the union of the rules is used.
• ZStack Cloud supports CPU overcommitment. Therefore, the number of vCPUs can be greater than that of pCPUs. However, if the number of vCPUs specified in the CPU pinning rule is greater than that of pCPUs, the VM performance will be affected. Therefore, CPU overcommitment is not recommended.
• You can modify the CPU pinning rules on the Overview tab page of a VM details page. The modification takes effect after you restart the VM instance.
• Cloning or migrating a VM instance also copies the CPU pinging rules, while creating a VM image or performing VM backup does not.
• When you create a VM instance, the Cloud firstly checks the VM scheduling policies and host allocation strategies, and then checks the CPU pinning rules.
• When you power off a VM instance and modify its instance offering by reducing the number of the its vCPUs, the CPU pinning rules related to the reduced vCPUs no longer take effect. In this case, we recommend that you modify or delete these rules.

• SSH Key: If you select this method, set the following parameters:
  • SSH Key: Select an SSH key to enable the root user to SSH in to the VM instance without entering a password when the VM instance is running.
  Note:

  • Make sure that the Cloud-Init is installed in your VM image.
  • If your VM image does not have the Cloud-Init installed, the SSH key attached to the VM instance cannot work properly. In this scenario, you can try one of the following methods to make the SSH key work:
    • Install the Cloud-Init for the VM instance after the creation and reboot it to make the SSH key work.
    • Install the Guesttools (or the QGA of 2.5 or latter version) for the VM instance after the creation. Detach the SSH key from the VM instance first and re-attach it. Then, the SSH key can work properly.
• Password: If you select this method, set the following parameters:
  • Username: The default username is root.
  • Password: After a root password is injected to your VM instance, you can SSH in to the VM instance by entering a password when the VM instance is running.
  Note:

  • Enter letters, digits, or special characters. Supported special characters include -`=[];',./~!@#$%^&*()_+|{}:"<>?.
  • Before you set a password, make sure that Cloud-Init is installed for the VM image.
  • If a VM instance runs CentOS, you can install Cloud-Init by running the yum install cloud-init command.
  • After you set the password, do not set it again via User Data to avoid conflicts of duplicated operations.
  • After you set the password, the password is displayed on User Data of the VM details page. Make sure that your password is well protected.
  • You can enable VM Password Strength in Global Setting to set password strength.

• Username: The default username is administrator.
• Password: After a password is injected to your VM instance, you can log in to the VM instance by using the password when the VM instance is running.

Note:

• Enter letters, digits, or special characters. Supported special characters include -`=[];',./~!@#$%^&*()_+|{}:"<>?.
• Before you set a login password, make sure that Cloudbase-Init is installed for the VM image. Recommended Cloudbase-Init version: 0.9.11.
• For more information about how to install Cloudbase-Init, see Cloudbase Documentation.
• After you set the password, do not set the it again via User Data to avoid conflicts of duplicated operations.
• After you set the password, the password is displayed on User Data of the VM details page. Make sure that your password is well protected.
• You can enable VM Password Strength in Global Setting to set password strength.

• Before you import user data, make sure that both the user data network service and DHCP network service work as expected.
• By default, the user data network service and DHCP network service in the flat network and VPC network environments are enabled.
• For Linux VM instances:
  • Before you inject User Data, make sure that Cloud-Init is installed for the VM image. Recommended Cloud-Init versions: 0.7.9, 17.1, 19.4, 19.4, and later.
  • If you set a password in SSH Login Method > Password, do not set a password again in User Data. Otherwise, a conflict may occur.
  • If you set the VM hostname both the Hostname input box and User Data. The value in the User Data takes effect.
  • If you create an Linux VM instance by using an image with Cloud-Init installed, you must import User Data. Otherwise, the Cloud-Init task will wait until timeout.
• For Windows VM instances:
  • Before you import User Data, make sure that Cloudbase-Init is installed for the VM image. Recommended Cloudbase-Init version: 0.9.11.
  • We recommend that you do not modify the default username in Cloudbase-Init. If the Cloudbase-Init default user name is the same as the VM OS username, Cloudbase-Init generates a random password for this OS user and overwrites the password you set.
  • For more information about how to install Cloudbase-Init, see the Cloudbase Documentation.
  • If you set the VM password in Login Method, do not set it again in User Data. Otherwise, a conflict may occur.
  • Do not set a hostname for a Windows VM instance in UserData during the creation. If you set one, this hostname does not take effect after the creation. To set the hostname for a Windows VM instance, finish the creation and install a GuestTools for the VM instance first.
  • If you create a Windows VM instance by using an image with Cloudbase-Init installed, you must import User Data. Otherwise, the Cloudbase-Init task will wait until timeout.
• After you set the login password in User Data, a clear-text password is displayed on User Data of the VM details page. Remember to secure your password.
• When you use User Data, note that: If an L2 network has multiple L3 networks within the same CIDR, only the user data of the first L3 network works. This will make the internal monitoring of the VM instance fail to work properly.
• You cannot use User Data if the VM instance uses an SR-IOV VF NIC.
• The following is an example of importing User Data to a Linux VM instance.

  #cloud-config
  users:
   - name: test
     shell: /bin/bash
     groups: users
     sudo: ['ALL=(ALL) NOPASSWD:ALL']
     ssh-authorized-keys:
         - ssh-rsa AAAAB3NzaC1LXCJfjroD1lT root@10-0-0-18
  bootcmd:
   - mkdir /tmp/temp 
  write_files:
   - path: /tmp/ZStack_config
     content: |
         Hello,world!
     permissions: '0755'
  fqdn: Perf-test
  disable_root: false
  ssh_pwauth: yes
  chpasswd:
    list: |
        root:word
    expire: False
  runcmd:
   - echo ls -l / >/root/list.sh

  The preceding script can do the followings:

  1. Create a user named test and use ssh-key when a VM instance is created.
  2. Write the /etc/hosts file when the VM instance is started, create a directory named /tmp/temp, create a file, and write data to the file.
  3. Set a hostname, enable the root user, enable SSH login with password, and change the root password.
  4. Run the echo ls -l / command.
• The following is an example of importing User Data to a Windows VM instance.

  #cloud-config
  write_files:
     -   encoding: b64
         content: NDI=
         path: C:\b64
         permissions: '0644'
     -   encoding: base64
         content: NDI=
         path: C:\b64_1
         permissions: '0644'
     -   encoding: gzip
         content: !!binary |
             H4sIAGUfoFQC/zMxAgCIsCQyAgAAAA==
         path: C:\gzip
         permissions: '0644'

  The preceding script can do the following:

  • Create b64, b64_1, and gzip files under C drive when the VM instance is started.

• Enter letters, digits, and special characters. Supported special characters include -`=[];',./~!@#$%^&*()_+|{}:"<>?.
• If you select the VNC mode, you can enable VNC Console Password Strength in Global Setting to set password strength.
• If you set a console password, you need to enter the password when you launch the console.

• Image BIOS modes include Legacy and UEFI. A VM instance inherits the BIOS mode from its image.
• You need to get the corresponding image ready, and select a proper BIOS mode for it.
• You can change the BIOS mode on the VM details page. Exercise with caution when you make any changes. The VM instance may fail to work properly because of an incompatible BIOS mode. After you change the BIOS mode, restart the VM instance to make the change take effect.
• The Legacy mode is recommended when you create a VM instance. If you want to use the UEFI mode, we recommend that you select the corresponding image from the following list of operating system versions.

  OS	BIOS Mode	Supported Versions
  Windows	UEFI	Windows 8 or later
  	UEFI	Windows 7 Windows Server 2008 R2
  Linux	UEFI	CentOS 7.2 CentOS 7.3 CentOS 7.4 or later

• By default, the console mode of a newly created VM instance is consistent with value of the global setting Console Mode.
• You can modify the console mode for a VM instance individually when you creating it or after the creation. Then, the global setting mentioned above does not take effect on this VM instance.
• If you modify the console mode for a VM instance, restart the VM instance to make the new mode take effect.

• By default, the console password must be 6~8 characters in length and can contain letters, digits, and the following special characters: -`=[];',./~!@#$%^&*()_+|{}:"<>?
• You can adjust the password requirement through the global setting VNC Console Password Strength Policy.
• If you set a console password for a VM instance, you must use the correct password to enter the VM console. If you do not set a console password or cancel the password, you can enter the VM console directly by clicking Launch Console.
• If there is a password set for the console, after changing the password, the new password will take effect immediately.
• If a new password is added for the console or the console password is removed, you can restart the VM instance(s) for the changes to take effect.

• Internal monitoring agent: An agent responsible for obtaining VM internal monitoring data, including CPU, memory, and disk capacity data.
• QEMU Guest Agent (QGA)：A program used for interactions between VM instances and enabling features like configuration deployment, configuration reading, and automated O&M.
  Note: Linux operating systems typically come with built-in VirtIO drivers and Cloud-Init, so the GuestTools for Linux VM instances does not include these two components separately. If your Linux operating system does not include VirtIO and Cloud-Init, you can install them through other methods.

• VirtIO driver: A collection of drivers to improve the VM performance.
  • SCSI controller: used to improve the disk performance of a Windows VM instance in a virtualization environment.
  • PCI simple communications controller: used to realize the communications between a Windows VM instance and the underlying KVM virtualization.
  • PCI device: used to realize the balloon memory scaling.
  • Ethernet adapter: used to improve the network performance of a Windows VM instance in a virtualization environment.
• Cloudbase-Init: A component helps realize the User Data import and other customized functions.

1. Click Install GuestTools on the VM details page.
2. Attach an ISO of VM GuestTools.
3. Run the following installation commands on the VM instance console:

   # Create a mount point.
   mkdir /mnt/cdrom
   # Mount the CND-ROM image.
   mount /dev/cdrom /mnt/cdrom
   # 
   cd /mnt/cdrom/
   bash ./zs-tools-install.sh
   # Unmount the CD-ROM image.(Optional)
   cd ~
   umount /mnt/cdrom

   Note:

   • You can copy the commands by one click from the VM details page to the VM instance console.
   • Before you install GuestTools, make sure that you have installed the appropriate Linux command tools, such as tar, wget, and curl.
   • For VM instance of OpenEuler OS, you need to disable selinux after installing GuestTools, otherwise it will affect the QGA function.

   
   
   
   
   

1. Click Install GuestTools on the VM details page.
2. Attach an ISO of VM GuestTools.
3. Launch the VM console and install GuestTools as prompted.

   
   
   

1. Click Install GuestTools on the VM details page.
2. Run the following installation commands on the VM instance console:

   curl http://169.254.169.254/vm-tools.sh -o vm-tools.sh && bash -x ./vm-tools.sh

   Note:

   • In the command above, 169.254.169.254 is the IP address of the User Data server.
   • Before you install GuestTools, make sure that you installed the bash, security/libgcrypt, and curl tools.

   
   
   
   
   

• Select All: Displays the utilization of all CPUs and the average CPU utilization. The average utilization and utilization of different CPUs are displayed in different colors.
• Average: Displays the average of all CPUs of a VM instance.
• Specified CPU: Displays the utilization of a specified CPU. You can specify one or more CPU. The utilization of different CPUs are displayed in different colors.

• Total traffic: The total amount of data sent or received by a NIC.
• Average data rate: The average amount of data sent or received per second by a NIC.
• Maximum data rate: The maximum value of the average data rate. The average data rate is calculated every 5 minutes.
• 95th percentile: This value is calculated by arranging the average data rate samples collected every 5 minutes in the selected time range in descending order, and removing all values greater than the highest 5% of the data.

• NIC data rate: The number of data bits sent or received by a NIC per second.
• NIC packet rate: The number of data packets sent or received by a NIC per second.

• If the default network of the VM instance has the DHCP service enabled, the monitoring data is pushed by the DHCP service.
• If the default network of the VM instance has the DHCP service disabled or the DHCP services fails, the monitoring data is then pushed by the QGA. After the DHCP service of the default network is recovered, the monitoring data is switched to be pushed by the DHCP service again.

• CPU Occupancy Rate (System Process): The percentage of CPUs occupied by the current kernel space of the VM instance.
• CPU Occupancy Rate (User Process): The percentage of CPUs occupied by the current user process of the VM instance.
• CPU Occupancy Rate (Waiting): The percentage of CPUs used for waiting for the I/O operations of the VM instance.
• CPU Idle Rate: The percentage of idle CPUs of the VM instance.
• CPU Utilization: The percentage of used CPUs of the VM instance.

• Select All: Display the average, real-time utilization of all CPUs of a VM instance and the utilization of a single CPU.
• Average: Display the average, real-time utilization of all CPUs of a VM instance.
• Single CPU: Display the real-time utilization of a single CPU, such as CPU No. 1, No.2, and No. 3.

• used: The used memory of a VM instance.
• available: The available memory of a VM instance.
• free: The idle memory of a VM instance.
• total: The total memory of a VM instance.
• freeutilization: The percentage of idle memory of a VM instance.
• usedutilization: The percentage of the used memory of a VM instance.

Note: For Windows 7 and Windows Server 2012 VM instances with Memory Reclaim enabled, the used and total memory displayed here contain the memory that has been reclaimed to the host.

• Disk Utilization: The percentage of used disk of a VM instance.
• Disk Used Capacity: The used capacity of the VM disk.
• Disk Idle Rate: The percentage of idle disk capacity of a VM instance.
• Disk Idle Capacity: The idle disk capacity of a VM instance.

• Select All: Display the real-time status of all disk partitions.
• Single disk partition: Display the real-time read/write status of a single disk partition, such as the vda partition.

• By NUMA Topology: This method indicates that you pin vCPUs to pCPUs based on the host pNUMA topology. You can implement Manual Pinning or Smart Pinning.
  • Manual Pinning allows you to customize vCPU-pCPU binding for all vCPUs.
  • Smart Pinning assigns a vCPU of a VM instance to a pCPU on a host pNUMA node by descending order of the pNUMA node ID. If all pCPUs on the pNUMA node are pinned, a pCPU on the next pNUMA node is to be pinned. If all pCPUs of the host are pinned but some vCPUs are not assigned to pCPUs, each vCPU is assigned to a pCPU starting again from the pNUMA node with the maximum node ID.
  • You can pin a vCPU to one or more pCPUs or pin one or more vCPUs to a pCPU.
  • The utilization of each pCPU in the past 15 minutes is displayed. You can pin a vCPU to the optimal pCPU based on the utilization.
• By Entry:
  • You can click Add CPU Pinning to specify more CPU pinning rules.
  • Enter the vCPU range in the left text box and the pCPU range in the right. Note that the - symbol indicates the value range, while the ^ symbol indicates that a value is not included. If you specify multiple entries in a rule, separate each entry by using a comma (,). The following are some examples:
    • 0-2 indicates CPU 0, CPU 1, and CPU 2.
    • ^2 indicates that CPU 2 is not included.
    • 0-2,^2 indicates CPU 0 and CPU 1.
    • 1-7,^2,^3,^4,10 indicates CPU 1, CPU 5, CPU 6, CPU 7, and CPU 10.

Note:

• ZStack Cloud supports fixed vCPU performance on VM instances.
• All vCPUs of the VM instance must be pinned to pCPUs of the host.
• The pCPUs pinned by a vCPU must belong to the same pNUMA node.
• The vCPU range depends on the selected instance offering of the VM instance, while the pCPU range depends on the number of pCPUs of the host.
• If a vCPU has multiple CPU pinning rules, the union of the rules is used.
• ZStack Cloud supports CPU overcommitment. Therefore, the number of vCPUs can be greater than that of pCPUs. However, if the number of vCPUs specified in the CPU pinning rule is greater than that of pCPUs, the VM performance will be affected. Therefore, CPU overcommitment is not recommended.

• The vCPUs that pin pCPUs residing on the same pNUMA node form a vNUMA node. Therefore, each vNUMA node is associated with a pNUMA node.
• Total Memory: The total amount of local memory that a vCPU can access from the vNUMA node where it resides.

• All pNUMA nodes that associate with the VM instance are displayed.
• Total Memory: The total amount of local memory that a pCPU can access from the pNUMA node where it resides.
• Free Memory: The amount of free local memory that a pCPU can access from the pNUMA node where it resides.
• Both the total memory and free memory are measured based on physical hardware resources.

Database services such as Oracle and SQL Server and compiling services need to frequently access memory. This requires high performance and fast response of the system. vNUMA enables the services to primarily access local memory and thus greatly reduces latencies and improves computing performance of the system.

CPU-intensive applications have high demand on CPU resources. If many small CPU-intensive applications are running on a VM instance, frequent context switching may occur. vNUMA resolves this issue by assigning vCPUs to specific pCPUs and thus improves system performance.

• • By NUMA Topology: This method indicates that you pin vCPUs to pCPUs based on the host pNUMA topology. You can implement Manual Pinning or Smart Pinning.
    • Manual Pinning allows you to customize vCPU-pCPU binding for all vCPUs.
    • Smart Pinning assigns a vCPU of a VM instance to a pCPU on a host pNUMA node by descending order of the pNUMA node ID. If all pCPUs on the pNUMA node are pinned, a pCPU on the next pNUMA node is to be pinned. If all pCPUs of the host are pinned but some vCPUs are not assigned to pCPUs, each vCPU is assigned to a pCPU starting again from the pNUMA node with the maximum node ID.
    • You can pin a vCPU to one or more pCPUs or pin one or more vCPUs to a pCPU.
    • The utilization of each pCPU in the past 15 minutes is displayed. You can pin a vCPU to the optimal pCPU based on the utilization.
  • By Entry:
    • You can click Add CPU Pinning to specify more CPU pinning rules.
    • Enter the vCPU range in the left text box and the pCPU range in the right. Note that the - symbol indicates the value range, while the ^ symbol indicates that a value is not included. If you specify multiple entries in a rule, separate each entry by using a comma (,). The following are some examples:
      • 0-2 indicates CPU 0, CPU 1, and CPU 2.
      • ^2 indicates that CPU 2 is not included.
      • 0-2,^2 indicates CPU 0 and CPU 1.
      • 1-7,^2,^3,^4,10 indicates CPU 1, CPU 5, CPU 6, CPU 7, and CPU 10.

Note:

• The vCPU range depends on the selected instance offering of the VM instance, while the pCPU range depends on the number of pCPUs of the host.
• If a vCPU has multiple CPU pinning rules, the union of the rules is used.
• ZStack Cloud supports CPU overcommitment. Therefore, the number of vCPUs can be greater than that of pCPUs. However, if the number of vCPUs specified in the CPU pinning rule is greater than that of pCPUs, the VM performance will be affected. Therefore, CPU overcommitment is not recommended.

CPU-intensive applications have high demand on CPU resources. If many small CPU-intensive applications are running on a VM instance, frequent context switching may occur. CPU pinning resolves this issue by assigning vCPUs to specific pCPUs and thus improves system performance.

If many applications are using one or more specific pCPUs, you can configure CPU pinning to adjust CPU usage in a more balanced way.

If many VM instances that run different applications are running on a host, context switching may occur. EmulatorPin assigns main QEMU threads of a VM instance to specific pCPUs. This isolates the main threads of VM instances and thus ensures relative stability of the system.