3.1. Compute Resource¶
Compute Resource is a service that provides virtual equipment (Compute Resources) by combining CPUs, Memory, and Disks to create Virtual Machines. Compute Resources are provided by virtualizing physical servers and storage devices shared by multiple users.
Use the Customer Portal to create, change, or delete a Virtual Machine.
3.1.1. Available Features¶
You can use the following features in Compute Resource.
| Feature | Overview | |
| 1 | Provision of Compute Resource Pools | A feature that uses the Compute Resources (CPU/Memory/Disk) to create Virtual Machines. You can create multiple machines. |
| 2 | Features for controlling Compute Resource Pools | From the Customer Portal, you can perform the following actions for Compute Resource Pools.
|
Note
- The infrastructure for Compute Resources is comprised of HA (High Availability) clusters and storage devices that have spare physical servers. If a failure is detected on a physical server that contains Compute Resources, the server is automatically replaced by a standby server.
- You can select Compute Resources that offer the appropriate performance level (Guaranteed, Premium, Standard) for your intended use.
3.1.2. Provision of Compute Resource Pools¶
You can create and use multiple Compute Resource Pools (CPUs/Memory/Disk) to create a Virtual Machine.
Use the Customer Portal to add, delete, and change Compute Resource Pools.
Important
- When using multiple Data Centers, there must be a Compute Resource Pool for each Data Center.
- Compute Resources (CPU/Memory/Disk) cannot be assigned to multiple Compute Resource Pools.
You can add or reduce the resources handled by one Compute Resource Pool within the ranges shown below.
| Resource | Lower Limit | Upper Limit | Application Unit |
| CPU | 1 GHz | 48 GHz | 1 GHz |
| Memory | 1 GB | 144 GB | 1 GB |
| Disk | 50 GB | 4,000 GB | 1 GB |
Compute Resource Pools are comprised of two types of classes: The Compute Class (CPU/Memory) and the storage class (Disks). Each of these is separated into two types of service classes (Premium and Standard) with different levels of performance. You can select the class that is appropriate for your intended use.
Important
- Select the service class when creating the Compute Resource Pool. You cannot change the service class after the Compute Resource Pool has been created.
| Classes | Resource | Service Class | Details |
|---|---|---|---|
| Compute Class | CPU Memory | Guaranteed | The CPU resource and Memory resource values for which you applied are guaranteed.SLA is applicable for this component. |
| Premium | The CPU resource and Memory resource values for which you applied are guaranteed. | ||
| Standard | The CPU resource and Memory resource values for which you applied are provided on a best effort basis. | ||
| Storage Class | Disk | Premium | High-speed Disk performance is provided. |
| Standard | Standard Disk performance is provided. |
*Guaranted Class service is provided on shared basis with enough resoruces. But customer should note that CPU and Memory performance are not “guaranteed” when the basis usage is too high. (For example the Large-scale batched processing has done.) So we recommend using Compute Resource (Dedicated Device) service for the system which requires real time enormous performance. |
Compute Classes
The differences between compute service classes (Premium or Standard) are shown below.
Compute Resources are comprised of storage devices and HA clusters that have more than one of the following two types of physical servers.
- Regular servers
- Standby servers (spare physical servers used for failure recovery)
When a failure is detected on a regular server, the HA Cluster feature automatically switches to the resources on a standby server (automatically recovers).
Important
- The HA Cluster feature does not detect any failures and perform an automatic recovery on a Virtual Machine that you have created.
- The HA Cluster feature does not guarantee the recovery of a Guest OS or applications running on a Guest OS, on a Virtual Machine that you have created.
When a failure is detected on a regular server, the Virtual Machine restarts on a standby server. The Virtual Machine that you created may temporarily stop until it restarts on the standby server.
As a result, if you have created a redundant configuration between multiple Virtual Machines but you have added the Virtual Machines to the same Compute Resource Pool, the redundant configuration may not behave as expected.
Zones are used to deal with this problem.
A zone is a group of physical equipment (physical servers and storage devices) that accommodate a Compute Resource Pool. You can choose either Zone A or Zone B for each Compute Resource Pool.
Virtual machines created from Compute Resource Pools with different zones run on different physical equipment, as shown below.
Example: When zones are set on Compute Resource Pools 1 to 3
| Compute Resource Pool | Zone | Virtual Machine | Physical Equipment Running the Virtual Machine |
|---|---|---|---|
| Compute Resource Pool 1 | Zone A | Virtual Machine i | Physical Equipment A |
| Virtual Machine ii | Physical Equipment A | ||
| Virtual Machine iii | Physical Equipment A | ||
| Compute Resource Pool 2 | Zone A | Virtual Machine | Physical Equipment A |
| Compute Resource Pool 3 | Zone B | Virtual Machine | Physical Equipment B |
Note
- For information on Data Centers that offer zones, refer to “1.3.2 Available Data Centers”.
- Zone function provides the availability of the physical serve that Virtual Machine would run. It does not provide the availability for Network devices.
3.1.3. Features for Controlling Compute Resource Pools¶
From the Customer Portal, you can perform the following actions for Compute Resource Pools.
| Feature | Overview |
| Add/reduce resources | A feature for adding and reducing the three types of resources (CPU/Memory/Disk) in a Compute Resource Pool. |
| Assign resources to a Virtual Machine | A feature for assigning Compute Resources (CPU/Memory/
Disk) to a Virtual Machine created in a Compute Resource Pool.
|
| Add or delete a Compute Resource Pool | A feature for adding or deleting a Compute Resource Pool. |
3.1.4. vApp Feature¶
vApp is a new feature that can be seen on Customer Portal ver2.0. vApp is a container for Virtual Machines which is managed by VMware. All functional characteristics of vApp is currently not supported in Enterprise Cloud. vApp for Enterprise Cloud can only support one single Virtual Machine.
3.1.5. Assigning Resources to a Virtual Machine¶
Create a Virtual Machine by assigning resources in a Compute Resource Pool (CPUs/Memory/Disk) to the Virtual Machine. The amount of resources that can be assigned to a Virtual Machine is different with Customer Portal ver1.0 and Customer Portal ver2.0.
You can also add or reduce resources for the Virtual Machine once you have created it.
Note
- The number of Virtual Machines that you can create depends on the number of contracted resources and the number of private IP addresses that can be used on a Server Segment. IP addresses are used for vFirewall, vLoad Balancer, Service Interconnectivity, and Virtual Machines. You can verify usage in the portal.
Virtual machines are made up of six components (vCPU/Memory/Disk/vNICs/Virtual CD/DVD drives/Guest OS).
Resources that can be assigned to a Virtual Machine (Customer Portal ver2.0)
*The amount of resources that can be assigned to Virtual Machine differ according to the Compute Class.
* The amount of vCPU, and Disk Capacity that could be assigned to each Virtual Machine differ depending on the Compute Class. The total disk capacity that could be assigned must be the amount which subtracts the memory capacity assigned to Virtual Machine from the leftover disk capacity of Compute Resource Pool.
Note
- If the leftover of Storage Resource for Compute Resource Pool is 3,500GB, and 128GB memory is being mounted, the maximum of total disk capacity is 3,372GB (= 3,500-128).
A vCPU is virtual CPU hardware that makes up a Virtual Machine.
From the Compute Resource Pool, you can specify the number of vCPUs and assign it to a Virtual Machine.
How many can be assigned?
The quantities of vCPUs that can be assigned to one Virtual Machine are shown below. Customer Portal ver2.0
| Service Menu | Compute Class | Min | Max | Step |
|---|---|---|---|---|
| Compute Resource (Shared Device) | Guaranteed | 1 | 32 | 1 |
| Premium | 1 | 8 | 1 | |
| Standard | 1 | 8 | 1 |
Note
- The number of vCPU is up to 8 if virtual hardware version is 7. Please mark this specification when Virtual Machine image is imported.
Socket
Socket of vCPU in some of the Customer Portal ver2.0 available Data Centers, number of cores per socket can be set. The combination of socket and core could be set within the amount of resource that can be assigned to each Virtual Machine.
Functional Availability at each Data Center
| JP | US | UK | FR | ES | SG | HK | MY | TH | |||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Yokohama No.1 | Kansai1 | Kansai1a | Saitama No.1 | Lundy | |||||||
| Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y |
Important
- You can only change the number of vCPUs when the Virtual Machine is powered off. Please do not change configuration in Partially Powered Off state.
vCPU processing capacity
The vCPU processing capacity is different for each Data Center. The processing capacity is the same as the physical processors listed in the table below.
| Data Center | Processor |
| Yokohama No.1 | 2010 Intel Xeon Processor (equivalent to a maximum of 2.5 GHz) |
| Kansai 1 | 2012 Intel Xeon Processor (equivalent to a maximum of 2.0 GHz) |
| Saitama No.1 | 2012 Intel Xeon Processor (equivalent to a maximum of 2.2GHz) |
| San Jose Lundy | 2012 Intel Xeon Processor (equivalent to a maximum of 2.2 GHz) |
| UK Hemel Hempstead2 | 2012 Intel Xeon Processor (equivalent to a maximum of 2.2 GHz) |
| France Paris2 | 2012 Intel Xeon Processor (equivalent to a maximum of 2.2 GHz) |
| Spain Madrid2 | 2012 Intel Xeon Processor (equivalent to a maximum of 2.2 GHz) |
| Singapore Serangoon | 2012 Intel Xeon Processor (equivalent to a maximum of 2.2 GHz) |
| Hong Kong Tai Po | 2009 Intel Xeon Processor (equivalent to a maximum of 2.7 GHz) |
| Thailand Bangna | 2012 Intel Xeon Processor (equivalent to a maximum of 2.0 GHz) |
| Malaysia Cyberjaya3 | 2012 Intel Xeon Processor (equivalent to a maximum of 2.2 GHz) |
Important
- The vCPU processing power varies depending on the following conditions. There is no guarantee that a vCPU will always operate at the maximum processing capacity.
- When the total vCPU processing capacity for Virtual Machines running in one Compute Resource Pool is more than the purchased Compute Resource Pool (CPU resources)
- The load condition of the Guest OS on the Virtual Machine
Understanding resource consumption
The CPU resources that are consumed from the Compute Resource Pool are the resources that are actually used by the Virtual Machine for computational processing.
Note
- If a vCPU assigned to a Virtual Machine is not running, CPU resources are not consumed from the Compute Resources.
- If computational processing by a vCPU reaches the CPU upper limit for the Compute Resource Pool for each Virtual Machine, the processing capacity is averaged between the Virtual Machines and operations continue.
Memory is virtual Memory hardware that makes up a Virtual Machine.
From the Compute Resource Pool, you can specify the Memory capacity and assign capacity to a Virtual Machine.
How many can be assigned?
You can add or reduce the Memory capacity that is assigned to one Virtual Machine within the ranges shown below. Customer Portal ver2.0
| Service Menu | Compute Class | Min | Max | Step |
|---|---|---|---|---|
| Compute Resource (Shared Device) | Guaranteed | 1 | 128 | 1 |
| Premium | 1 | 32 | 1 | |
| Standard | 1 | 32 | 1 |
Important
- You can only change the Memory capacity when the Virtual Machine is powered off. Please do not change configuration in Partially Powered Off state.
Understanding resource consumption
The capacity totals below are consumed from the Compute Resource Pool.
- Total Memory capacity set for Virtual Machines that are running
- Memory resources for virtualization overheads
For information regarding overheads, refer to “Default Gateway“
When vFirewall/INA is not set as Default Gateway, it’s necessary to set specific Static Route additionally in Guest OS. For details, please check Server Segment section.
Snapshot
A snapshot is reproduction of vApp (virtual machine) just as it was when Customer took the snapshot. The snapshot includes the state of the data on all virtual machine disks at a given point in time. Customer can take or restore it by Customer Portal or API.
Note
- The data is different form Image backup or File backup, so data is not kept as physical data. Data does nothing but be kept logically.
Functional Availability at each Data Center
| JP | US | UK | FR | ES | SG | HK | MY | TH | |||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Yokohama No.1 | Kansai1 | Kansai1a | Saitama No.1 | ||||||||
| N | N | N | N | N | Y | Y | Y | Y | Y | Y | Y |
*After April, 2016 the function will become effective in sequence.
One generation Snapshot is kept.
Important
- When Snapshot will be taken again during execution, it is overwritten by new data.
Retention period is two days (48 hours). Snapshot data will be deleted when the retention period expires.
Important
- Notification about deletion is not send, so we recommend executing deletion by Customer before retention period respires.
Stored data in Snapshot listed in the table below.
| Item | Information | Detail |
|---|---|---|
| vApp | Friendly Name | - |
| Explanation | - | |
| Virtual Machine | Friendly Name | - |
| Explanation | - | |
| vCPU | - | |
| Memory | Setting of Memory Option is needed. | |
| Disk | Including Data in disk | |
| vNIC | Network settings in both Customer Portal and Guest OS | |
Other Devices CD/DVD driveand so on |
Device information only |
Disk Resource in Compute Resource is needed for Snapshot. All disk volume attached Virtual Machine and assigned Memory (if memory option is set) is consumed until Snapshot deletion.
| Item | Outline |
| Take Snapshot | Taking Snapshot on time when it was operated. |
| Restore Snapshot | Restoring vApp (Virtual Machine) |
| Delete Snapshot | Deleting Snapshot |
Take Snapshot
Snapshot options are listed below.
| Option | Outline |
| Memory | If this option is used, a dump of the internal state of the virtual machine (basically a memory dump) is included in the snapshot. |
| Quiesce | If this option is used, the file system quiesce on Guest OS is taken. |
Important
- To use these options virtual machine must be powered on when the snapshot is taken.
- Must have the most up-to-date VMware tools installed, and must be enabled to use Quiesce option
- Quiesce option is not guaranteed success. The fault will be happened because of the user-setting of Guest OS or applications, so please test before actual operation.
Restore Snapshot
The snapshot includes the state of the virtual machine power state. So Virtual Machine is restored the same state.
Delete Snapshot
Customer can delete snapshot. Even if deletion is executed, real Virtual Machine itself is not be lost.
Important
- To execute deletion the state of the virtual machine power state must be PoweredOff or PoweredOn.
- It sometimes even takes several hours for completion if Customer does the operation it takes load on disk (read and write) in Virtual Machine.
- During Snapshot taking, the functions listed below are not available.
| Name of Menu/Feature | create/execute | Display | Change | Delete | ||
|---|---|---|---|---|---|---|
| Private Catalog | Template | N | - | - | - | |
| Resource | vCPU | - | Y | N | - | |
| Memory | - | Y | N | - | ||
| Virtual Machine/vApp | Number of disks | N | Y | - | N | |
| Disk Capacity | - | Y | N(extension) | - | ||
| vNIC(Select Server Segment) | N | Y | - | - | ||
| ISOImage Mount Feature | N | - | - | - | ||
| Set Guest Customization Enabled | N | - | - | - | ||
| Windows OS SID Modification Feature | N | - | - | - | ||
| Friendlyname | - | Y | N | - | ||
| Explanation | - | Y | N | - | ||
When Virtual Machine state is PoweredOn, the performance of the Disk I/O of the Virtual Server might be reduced or stopped tens of seconds if customer take or delete Snapshot.
Important
- It is recommended to take snapshot in the state of PoweredOff if the affection cannot be estimated.
- Please test this function and confirm the influence to the system before actual operation.
Important
- The available Memory capacity varies depending on the following situations. There is no guarantee that the maximum Memory capacity will be always available.
- The usage status of Memory resources for which you have applied
- The load condition of the Guest OS on the Virtual Machine
Note
- When the Memory resources consumed on each Virtual Machine reach the upper limit of Memory for the Compute Resource Pool, Memory in the swap regions of the Disk resources may be activated.
A Disk is a virtual storage device that makes up a Virtual Machine.
From the Compute Resource Pool, you can specify the Disk capacity and assign capacity to a Virtual Machine.
There are two types of Disks: a root Disk and a data Disk.
| Disk | Description |
| Root Disk | The Disk that stores the Guest OS.
There is always one root Disk created for one Virtual Machine.
|
| Data Disk | The Disk that stores data.
You can connect multiple Disks for one Virtual Machine.
|
Important
- If a Virtual Machine is deleted, the root Disk and data Disks are deleted at the same time.
- The data from a deleted Disk is erased according to the appropriate method specified by NTT Communications. A data erasure certificate is not issued.
- You cannot remove (detach) a data Disk that is connected to a Virtual Machine and connect (attach) it to another Virtual Machine.
Note
- You can add and delete data Disks and expand the Disk capacity from the Customer Portal, regardless of whether the Virtual Machine is powered on or off. But please do not change in Partially Powered Off state.
- If you add or delete a data Disk or expand the Disk capacity while the Virtual Machine is powered on, the Disk may not be recognized properly by the Guest OS. However, it will be recognized properly if the Guest OS is compatible with hot swap.
- The Disk capacity of the root Disk depends on the template that was selected when creating the Virtual Machine.
How many can be assigned?
You can add or reduce the Disk capacity and the number of data Disks connected to one Virtual Machine within the ranges shown below.
Customer Portal ver2.0
| Lower Limit | Upper Limit | Setting Unit | |
|---|---|---|---|
| Number of data Disks | 0 | 59 | 1 |
| Disk capacity | 1 GB | 2,047GB | 1 GB |
| 1 MB | 2,097,151 MB | 1 MB |
Important
- There is no limit for total disk capacity. However, the total disk capacity (no limit) + Memory Resource (different for each Compute Class) must be below the amount of space left in storage resource.
Understanding resource consumption
The capacity totals below are consumed from the Compute Resource Pool.
- Total Disk capacity assigned to a Virtual Machine
- Capacity of swap regions for each Virtual Machine (same capacity as the Memory capacity)
A vNIC is virtual network adapter hardware that makes up a Virtual Machine.
The Server Segment service provides an L2 connection to Server Segments in the same Data Center.
Important
- A separate application is required to use the Server Segment service.
- One of the assigned vNICs must be set as the representative vNIC (called the “Primary vNIC” below). Some of the initial settings for the Guest OS are affected by the primary vNIC selection. For details, refer to “Guest OS Customization”.
- Monitoring of Virtual Machine pings is performed for the primary vNIC.
Note
- You can specify settings for an L2 connection between a primary vNIC and a Server Segment only when creating a Virtual Machine or when the Virtual Machine is powered off. Specify the settings from the Customer Portal.
- You cannot connect multiple vNICs from the same Virtual Machine to one Server Segment.
How many can be assigned?
Eight vNICs can be used on one Virtual Machine. This cannot be changed.
Note
- You can assign IP addresses to vNICs when creating a Virtual Machine. You can also change the IP address that is assigned to a vNIC.
- The system can automatically assign an IP address to a vNIC. To use this option, select Auto Assign.
- The system can automatically assign the IP address to vNIC from the available IP addresses in the IP address block specified by the Server Segment. You can also set an IP address from the Customer Portal.
- Sub-interface settings other than the IP addresses assigned to vNICs are specified on the Guest OS. To change an IP address in the sub-interface settings, you must first register the IP address that you want to assign as a reserved IP.
A virtual CD/DVD drive is virtual CD/DVD-ROM drive hardware that makes up a Virtual Machine.
Important
- You can connect only one virtual CD/DVD drive to one Virtual Machine. The number of virtual CD/DVD drives cannot be changed.
Only Guest OSes that are supported by vCloud Director can be used with Virtual Machines.
|
|
Important
- Install and enable the latest VMware Tools in the Guest OS on the Virtual Machine. If you intentionally uninstall or disable VMware Tools, we cannot guarantee the correct operation of Compute Resources. We also may not be able to support your queries.
Guest OS Customization
Guest OS settings basically depend on the template. However, some settings are automatically changed after power on at the first time in following operation. This is referred to as “Guest OS Customization”.
1) After creating a Virtual Machine
2) After changing the Server Segment to which a vNIC connects
3) After changing the primary vNIC
4) After changing the IP address of the vNIC
Important
- The Virtual Machine automatically restarts when the Guest OS is customized. Do not log in to the Guest OS or operate the Virtual Machine until it has restarted. The Virtual Machine will operate in the state that it was in prior to customization of the Guest OS, until it restarts.
- Please do not operate Virtual Machine during Guest OS Customization. Usually, it takes about 30 minutes.
Settings that are changed when customizing the Guest OS
The Guest OS settings that are changed when customizing the Guest OS are shown below.
- Items that are changed automatically when turning the power on for the first time after creating a Virtual Machine.
| Item | Setting | Remarks |
| IP Address | A value specified by the user or by NTT Communications | Applies to all vNICs. |
| Net mask | The subnet mask of the Server Segment to which the vNIC connects | Applies to all vNICs. |
| Default gateway | A value specified by the user or by NTT Communications (*) | - |
| Primary DNS | A value specified by the user or by NTT Communications | - |
| Secondary DNS | A value specified by the user or by NTT Communications | - |
| DNS suffix | A value specified by the user or no value | - |
| S-ID | - | For Windows OS only, a Sysprep is performed and the S-ID is changed automatically. |
| root/Admin password | A value specified by NTT Communications | - |
| Host/computer name | A value specified by NTT Communications | - |
* The settings that are specified by NTT Communications are the IP addresses for the vFirewall/Integrated Network Appliance for the Server Segments to which the primary vNIC connects. However, the IP address that is set for Server Segments that do not connect to the vFirewall/Integrated Network Appliance is “the “broadcast address” of the IP address block for the Server Segment - 1.” For example, if the IP address block is “192.168.0.0/24,” the IP address that is “the “broadcast address” of the IP address block for the Server Segment - 1” will be “192.168.0.254.”
- Settings that are changed automatically when starting for the first time after changing the Server Segment to which the vNIC connects, the primary vNIC, or the vNIC IP address
| Item | Setting | Remarks |
| IP Address | A value specified by the user or by NTT Communications | Applies to the vNIC for which the destination Server Segment has changed. |
| Net mask | The subnet mask of the Server Segment to which the vNIC connects | Applies to the vNIC for which the destination Server Segment has changed. |
| Default gateway | A value specified by the user or by NTT Communications (*) | - |
| Primary DNS | A value specified by the user or by NTT Communications | - |
| Secondary DNS | A value specified by the user or by NTT Communications | - |
| DNS suffix | A value specified by the user or no value | - |
| Host/computer name | A value specified by NTT Communications | - |
* The settings that are specified by NTT Communications are the IP addresses for the vFirewall/Integrated Network Appliance for the Server Segments to which the primary vNIC connects. However, the IP address that is set for Server Segments that do not connect to the vFirewall/Integrated Network Appliance is “the “broadcast address” of the IP address block for the Server Segment - 1.” For example, if the IP address block is “192.168.0.0/24,” the IP address that is “the “broadcast address” of the IP address block for the Server Segment - 1” will be “192.168.0.254.”
Note
- The S-ID and root/Admin password does not change.
- Contents that are automatically changed at the initial start after restoring the Image Backup
| Item | Setting value | Remarks |
| Net Mask | Subnet mask of the server segment to which the vNIC is connected | Applies to all vNICs. |
| Gateway | Value specified by customer or NTT Communications *1 | - |
| Primary DNS | Value specified by customer or NTT Communications | - |
| Secondary DNS | Value specified by customer or NTT Communications | - |
| DNS suffix | Value specified by customer or no value | - |
| Host name/ Computer name | Value specified by NTT Communications | - |
*1 The values specified by NTT Communications are the IP addresses for the vFirewall/Integrated Network Appliance for the Server Segments to which the primary vNIC connects. However, the IP address that is set for Server Segments that do not connect to the vFirewall/Integrated Network Appliance is “the “broadcast address” of the IP address block for the Server Segment - 1.” For example, if the IP address block is “192.168.0.0/24,” the IP address that is “the “broadcast address” of the IP address block for the Server Segment - 1” will be “192.168.0.254.”
Note
- IP address, root/Admin password, mac address are restored with values upon backup. Other parameters are changed to the setting values described in the above table. Note that parameters which changed in Guest OS are not recovered.
- S-ID is not changed.
- Default Gateway
When vFirewall/INA is not set as Default Gateway, it’s necessary to set specific Static Route additionally in Guest OS. For details, please check Server Segment section.
3.1.6. Snapshot¶
A snapshot is reproduction of vApp (virtual machine) just as it was when Customer took the snapshot. The snapshot includes the state of the data on all virtual machine disks at a given point in time. Customer can take or restore it by Customer Portal or API.
Note
- The data is different form Image backup or File backup, so data is not kept as physical data. Data does nothing but be kept logically.
Functional Availability at each Data Center
| JP | US | UK | FR | ES | SG | HK | MY | TH | |||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Yokohama No.1 | Kansai1 | Kansai1a | Saitama No.1 | ||||||||
| N | N | N | N | N | Y | Y | Y | Y | Y | Y | Y |
*After April, 2016 the function will become effective in sequence.
One generation Snapshot is kept.
Important
- When Snapshot will be taken again during execution, it is overwritten by new data.
Retention period is two days (48 hours). Snapshot data will be deleted when the retention period expires.
Important
- Notification about deletion is not send, so we recommend executing deletion by Customer before retention period respires.
Stored data in Snapshot listed in the table below.
| Item | Information | Detail |
|---|---|---|
| vApp | Friendly Name | - |
| Explanation | - | |
| Virtual Machine | Friendly Name | - |
| Explanation | - | |
| vCPU | - | |
| Memory | Setting of Memory Option is needed. | |
| Disk | Including Data in disk | |
| vNIC | Network settings in both Customer Portal and Guest OS | |
Other Devices CD/DVD driveand so on |
Device information only |
Disk Resource in Compute Resource is needed for Snapshot. All disk volume attached Virtual Machine and assigned Memory (if memory option is set) is consumed until Snapshot deletion.
| Item | Outline |
| Take Snapshot | Taking Snapshot on time when it was operated. |
| Restore Snapshot | Restoring vApp(Virtual Machine) |
| Delete Snapshot | Deleting Snapshot |
Take Snapshot
Snapshot options are listed below.
| Option | Outline |
| Memory | If this option is used, a dump of the internal state of the virtual machine (basically a memory dump) is included in the snapshot. |
| Quiesce | If this option is used, the file system quiesce on Guest OS is taken. |
Important
- To use these options virtual machine must be powered on when the snapshot is taken.
- Must have the most up-to-date VMware tools installed, and must be enabled to use Quiesce option
- Quiesce option is not guaranteed success. The fault will be happened because of the user-setting of Guest OS or applications, so please test before actual operation.
Restore Snapshot
The snapshot includes the state of the virtual machine power state. So Virtual Machine is restored the same state.
Delete Snapshot
Customer can delete snapshot. Even if deletion is executed, real Virtual Machine itself is not be lost.
Important
- To execute deletion, the state of the virtual machine power state must be PoweredOff or PoweredOn.
- It sometimes even takes several hours for completion if Customer does the operation it takes load on disk (read and write) in Virtual Machine.
- During Snapshot taking, the functions listed below are not available.
| Name of Menu/Feature | create/execute | Display | Change | Delete | ||
|---|---|---|---|---|---|---|
| Private Catalog | Template | N | - | - | - | |
| Resource | vCPU | - | Y | N | - | |
| Memory | - | Y | N | - | ||
| Virtual Machine/vApp | Number of disks | N | Y | - | N | |
| Disk Capacity | - | Y | N（extension） | - | ||
| vNIC（Select Server Segment） | N | Y | - | - | ||
| ISOImage Mount Feature | N | - | - | - | ||
| Set Guest Customization Enabled | N | - | - | - | ||
| Windows OS SID Modification Feature | N | - | - | - | ||
| Friendlyname | - | Y | N | - | ||
| Explanation | - | Y | N | - | ||
When Virtual Machine state is PoweredOn, the performance of the Disk I/O of the Virtual Server might be reduced or stopped tens of seconds if customer take or delete Snapshot.
Important
- It is recommended to take snapshot in the state of PoweredOff if the affection cannot be estimated.
- Please test this function and confirm the influence to the system before actual operation.
3.1.7. Important Points¶
Resources Consumed by the Memory and Disk Overhead Regions In Connection With Server VirtualizationVirtual machines have four types of power states. The consumption of resources in the overhead regions for server virtualization depends on the power state. The overheads therefore need to be taken into account when designing the system (designing resources).
Each power state and the overhead regions required for each power state are shown in the table below.
The items marked with a “Y” are items that consume resources in overhead regions. For example, if the power state is Powered Off, resources from the overhead are not consumed for the CPU and Memory. On the other hand, the overhead portion consumes resources for the Disks.
| Power State | Meaning of Power State | CPU | Memory (*1) | Disk (*2) |
| Powered Off | The power for the Virtual Machine is off. | - | - | Y |
| Partially Powered Off | The power for the Virtual Machine is on but the Guest OS is stopped. | - | - | Y |
| Powered On | The power for the Virtual Machine is on. | Y | Y | Y |
| Suspended | The operation of the Virtual Machine has been stopped temporarily using the cloud infrastructure.
The suspend state and sleep state for the Guest OS is different to hibernation.
|
- | - | Y |
*1 The following overhead regions are required based on the number of vCPUs.
*2 The capacity of Disk resources consumed as the swap region is the same as the used Memory capacity.
Memory resource overheads (reference values*)
Important
*There are difference memory overhead size between initial PoweredOn status and steady status. *Memory will be consumed lager in initial PoweredOn status than in steady status, so please consider it and assign memory to Virtual Server. If being lack of overhead memory, Virtual Server cannot be PoweredOn.
Memory overhead in PoweredOn steady status. (reference values)
Memory
OH(GB)
|
Memory set on VM（GB） | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128 | 256 | 512 | ||
| vCPU | 1 | 0.11 | 0.13 | 0.16 | 0.23 | 0.37 | 0.64 | 1.19 | 2.29 | 4.49 | 8.87 |
| 2 | 0.13 | 0.15 | 0.18 | 0.25 | 0.39 | 0.66 | 1.21 | 2.31 | 4.51 | 8.90 | |
| 4 | 0.18 | 0.19 | 0.23 | 0.30 | 0.43 | 0.71 | 1.25 | 2.36 | 4.55 | 8.94 | |
| 8 | 0.26 | 0.28 | 0.32 | 0.38 | 0.52 | 0.80 | 1.35 | 2.44 | 4.64 | 9.04 | |
| 16 | 0.44 | 0.46 | 0.49 | 0.56 | 0.70 | 0.97 | 1.52 | 2.62 | 4.82 | 9.21 | |
| 32 | 0.79 | 0.81 | 0.84 | 0.91 | 1.05 | 1.32 | 1.88 | 2.97 | 5.17 | 9.56 | |
Temporary memory overhead in initial PoweredOn status. (reference values)
Memory
OH(GB)
|
Memory set on VM（GB） （GB） | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128 | 256 | 512 | ||
| vCPU | 1 | 0.11 | 0.13 | 0.17 | 0.25 | 0.41 | 0.73 | 1.34 | 2.6 | 5.12 | 10.16 |
| 2 | 0.13 | 0.16 | 0.2 | 0.29 | 0.48 | 0.85 | 1.55 | 3 | 5.89 | 11.69 | |
| 4 | 0.18 | 0.21 | 0.27 | 0.38 | 0.62 | 1.05 | 1.97 | 3.8 | 7.45 | 14.76 | |
| 8 | 0.27 | 0.31 | 0.39 | 0.56 | 0.89 | 1.51 | 2.81 | 5.4 | 10.58 | 20.95 | |
| 16 | 0.45 | 0.51 | 0.65 | 0.91 | 1.4 | 2.44 | 4.5 | 8.64 | 16.91 | 33.46 | |
| 32 | 0.81 | 0.92 | 1.12 | 1.57 | 2.47 | 4.26 | 7.85 | 15.03 | 29.38 | 58.09 | |
* Our test environment is shown below. This value will be changed according to user environment (Application, Operating System and so on).
- Guest OS: Red Hat Enterprise Linux 6.2 64-bit
- The number of CPU socket: 1-32
- CPU core per socket: 1
- Memory [GB]: 1-512
- Disk: default root disk only (The initial condition by which Virtual Machine was deployed)
- vNIC (E1000): 8
- Allocate one Server Segment IP address block to one Server Segment and specify the prefix length. Specify a prefix length of /29 to /24 for each Server Segment.
- NTT Communications manages the allocated IP address block for the Server Segment, and assigns the IP address selected from the IP address block to each device that connects to that Server Segment. For details, please check the description of features for each service.
- In the IP address block for the Server Segment, there are IP addresses blocks customer can’t specify or include (Non-duplicable).
Important
- For details about Non-duplicable IP Address blocks, refer to separate volume “Functional Description (IP Address)”.
- The IP address block for the Server Segment cannot be changed after it is allocated.
- If multiple Virtual Machines with the same role are created for one physical server and that physical server fails, the applications on those Virtual Machines may stop at the same time.
- You cannot select a physical server that runs a specific Virtual Machine.
- The network equipment and physical server interface provided by Compute Resource has redundancy. If the interface fails, it automatically switches from the regular interface to the standby interface. The Guest OS on the Virtual Machine and the applications that are running on the Guest OS may be affected when switching interfaces.
- If the zone is the same, resources may be kept on the same physical server or storage device, even if the service class (Premium or Standard) is different.
- In maintenance there is a possibility that Virtual Machines may move to another physical server using Live Migration function. Once it happens, Virtual Machine will be momentarily stopped, however there is no effect in general use of Guest OS and applications. And there are two possibilities that performance may fall and packet loss according to the loading state of the virtual server and applications.
- The performance of each resource may vary by Data Center.
- When changing Compute Resources, you need to create the Virtual Machines and configure the resource settings for Virtual Machines yourself. NTT Communications is not responsible for errors that occur as a result of these settings, such as abnormal operation of your applications.
Important
- When changing Compute Resources, we may ask you to create a new Compute Resource Pool to ensure that a stable service is provided, even if the compute resource that you are changing has not reached the resource upper limits.
- To use the Disk capacity expansion feature, you need to install and enable VMware Tools (Version 8.6.0 or higher) in the Guest OS on the Virtual Machine.
- The Disk capacity expansion feature cannot be used while a backup image is being obtained.
Important
- You cannot reduce the Disk capacity.
- You cannot change MAC addresses that have been set on virtual hardware such as vNIC.
- You cannot use your own MAC addresses that are not administered by NTT Communications.
- If we become aware that you have changed a MAC address or are using your own MAC address, we may stop that Virtual Machine without advance notice.
- When installing a Guest OS on a Virtual Machine, you need to verify the system requirements for the Guest OS (number of vCPUs, Memory capacity, Disk capacity, and so on), licenses, and terms of support with your Guest OS vendor yourself.
- When installing applications on a Guest OS, you need to verify the system requirements for the application (number of vCPUs, the CPU processing capacity of the vCPU, Memory capacity, number and capacity of Disks, number of vNICs, and so on), licenses, and terms of support with your application vendor yourself.
- When you install a Guest OS or application, NTT Communications is not responsible for checking or reporting whether operations can be guaranteed in your system configuration or whether there are any licensing issues.
- The Guest OS will recognize a vNIC as a NIC, even if it is not connected to a Server Segment. When changing the Guest OS network settings, do not disable a vNIC that has been recognized, even if you are not using that vNIC. If you do disable it, errors may occur in services such as Private Catalog and Image Backup.
- Compute Resource uses software that NTT Communications has licensed from VMWare, Inc.
- The VMware features provided in Compute Resource have been selected based on Compute Resource specifications. Not all VMware features are included.
- The following virtualization software is used in Compute Resource.
- VMware vSphere
- VMware vCloud Director
- Equivalent successor products
- New sale of Premium Compute and Standard Compute is suspended.