Skip to main content

Create and manage volume

info

This guide covers GPU and NPU instances. Video transcoding instances are provided on Virtual Machines, and for more information, please refer to Virtual Machine > Create instance.

The method of managing volumes and snapshots connected to virtual machine instances in GPU is as follows:

Create volume​

Create a new volume.

  1. Go to KakaoCloud Console > Beyond Compute Service > GPU.

  2. Select the Volume menu, then click on the [Create volume] button located at the top right.

  3. Enter the required information and click the [Create] button.

    이미지 Create volume

    CategoryDescription
    Volume NameEnter the name of the volume
    Volume Description (Optional)Enter a description for the volume
    TypeThe type of volume
    SizeThe size of the volume
    - Creation possible from 1~5,120GB
    Maximum IOPSThe maximum IOPS depending on the set volume size
    - 3,000 IOPS ~ 16,000 IOPS
    AZThe AZ location where the volume will be created
    Snapshot ScheduleNot used
    - The created volume does not use a snapshot schedule
    Used
    - The created volume automatically creates snapshots according to the set policy through the snapshot schedule
    EncryptionStored data is encrypted for security
    - Used as the default setting, cannot be changed

Manage volume​

View volume list​

View the list of operational volumes and basic information for each volume.

  1. Go to KakaoCloud Console > Beyond Compute Service > GPU.

  2. In the Volume menu, view the list of currently operational volumes. If needed, use various filter properties.

    CategoryDescription
    VolumeThe name and description of the volume
    StateThe status of the volume
    - For detailed volume status, refer to Volume state
    Attached InstanceThe connected instance and mount point
    - Clicking on the instance name redirects to the instance's detail page
    - Hovering over the UUID shows the instance ID
    TypeThe type of volume
    SizeThe size of the volume
    Snapshot ScheduleThe associated snapshot schedule
    AZAvailability zone where the volume is created
    CreatedThe creation date and time of the volume
    MoreExecute item-specific functions
    - Modify volume: Volume settings to apply to the instance
    - Attach/detach volume: Setting up and releasing the connection to the instance
    - Create snapshot: Creating a snapshot for the volume
    - Associate/Disassociate snapshot schedule: Setting up and releasing the connection to the snapshot schedule
    - Delete volume: Deleting the created volume

View volume details​

View and manage detailed information and snapshot information created from the Selected volume.

  1. Go to KakaoCloud Console > Beyond Compute Service > GPU.

  2. In the Volume menu, select the volume to view detailed information.

  3. On the volume's detail page, find the information.

    CategoryDescription
    ⓵      Volume name and description
    ⓶Clicking the [More] icon shows the full list of executable features                  
    - For the feature list, refer to volume more options
    ⓷Detailed information about the volume
    - Operating status: For detailed volume state, refer to Volume state
    - Size: The size of the volume
    - Type: The type of volume
    - Quick Execution: Quick execution feature
    ⓸The size of the volume
    Details tab

    View detailed information about the volume.

    1. Go to KakaoCloud Console > Beyond Compute Service > GPU.

    2. In the Volume menu, select the volume for detailed information.

    3. In the Detail tab, view the detailed information about the volume.

      CategoryDescription
      Volume IDThe unique ID of the volume
      - Click the icon to copy
      Volume NameThe name of the Selected volume
      CreatorThe user who created the volume
      - Click the link to view the list of instances created by the user
      CreatedThe creation date and time of the volume
      Attached InstanceThe name of the instance connected to the volume
      - Click the link to move to the detailed information of the instance
      Mount PointThe mount point of the volume
      Snapshot ScheduleThe snapshot schedule associated to the volume
      - Click the link to move to the snapshot schedule list
      AZAvailability zone where the volume was created
      TypeThe type of volume
    Snapshot tab

    You can view information about snapshots created from the volume. For more information, please refer to Create and manage snapshots/schedule.

    1. Go to KakaoCloud Console > Beyond Compute Service > GPU.

    2. In the Volume menu, select the volume to view snapshots.

    3. In the Snapshot tab, find the list of snapshots.

      CategoryDescription
      TypeThe type of snapshot
      - Full: Saves the entire volume as a snapshot
      - Incremental: Only saves changes from the last full snapshot
      SnapshotThe name and description of the created snapshot
      StateThe status of the snapshot
      SizeThe size of the snapshot
      Snapshot schedule- -: Snapshots created by the user or snapshots created through a snapshot schedule that was later deleted or disconnected
      - Snapshot schedule name: Snapshots automatically created according to the policy
      CreatedThe creation date of the snapshot
      - Hover the mouse to see the creation date and elapsed time
      MoreExecute item-specific functions
      - Modify snapshot: Change the name and description of the snapshot
      - Restore snapshot: Create a volume from the stored snapshot
      - Delete snapshot: Delete the snapshot

View volume by instance​

You can view the volume information connected to an instance.

  1. Go to KakaoCloud Console > Beyond Compute Service > GPU.
  2. In the Instance menu, select the instance to view volume information.
  3. In the instance's detail page, under the Volume menu, you can view the volume information of that instance.

Modify volume​

Modify the settings of a volume.

  1. Go to KakaoCloud Console > Beyond Compute Service > GPU.

  2. In the Volume menu, select the [More] icon > Modify volume.

    • When modifying the volume size, you can only enter values larger than the existing volume size in 1GB increments. After expanding the volume size, refer to Expand volume to update the size of the file system.

  3. Set up the volume and click the [Save] button.

Detach volume​

Detach a volume that is no longer in use.

info
  • The root volume being used as the OS area in an instance cannot be detached.
  • If you attempt to detach a volume while the instance is in the Active (running) state, you must first unmount the volume from the instance beforehand. For safe volume detachment, it is recommended to stop the instance and then proceed with the detachment.
  1. Go to KakaoCloud Console > Beyond Compute Service > GPU.
  2. In the Volume menu > volume list, select the [More] icon > Detach volume.
  3. Confirm the volume and instance to be detached, and click the [Detach] button.

Expand volume​

Expand the size of a volume.

info
  • After expanding the volume size, use file system-related commands to expand the file system size.
     ᄂ Start expanding the file system after the volume status changes to Extending.
  • After expanding the volume size, to apply the new volume QoS (Quality of Service), the following actions are required depending on the type of volume:
     ᄂ For root volumes, terminate (stop) the instance and then unshelve (start) the instance.
     ᄂ For additional volumes, detach the volume and then attach the volume.
  1. Go to KakaoCloud Console > Beyond Compute Service > GPU.
  2. Click the Volume menu, then select the [More] icon > Modify volume.
  3. Enter the expanded capacity in the Size field, and click the [Save] button.

Attach volume​

A newly created volume is in the Available state and is not attached to any instance. To use the volume properly, attach the volume to an instance through the console and follow the OS-specific guide to connect the volume as a device.

Attach volume from Console​

The method to connect a volume to an instance via KakaoCloud Console is as follows:

  1. Go to KakaoCloud Console > Beyond Compute Service > GPU.
  2. In the Volume menu, select the [More] icon > Attach volume.
  3. Enter the required information and click the [Save] button.

Attach volume on Linux​

In a Linux environment, access the instance via SSH to set up the file system and mount point. For details, please refer to Connect to Linux instance.

  1. Use the lsblk command to check whether the volume connected from the console is properly connected as a device.

    • e.g. Execute the lsblk command to confirm that the newly created 8G size volume is connected to /dev/vdb.
    lsblk command (check device connection)
    ubuntu@volume-attach-1:~$ lsblk
    NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
    vda 252:0 0 20G 0 disk /boot
    vdb 252:16 0 8G 0 disk

  2. Use the file -s command to check the file system type. If data is output, there is no file system on the device.

    file -s command (check file system type)
    ubuntu@volume-attach-1:~$ sudo file -s /dev/vdb
    /dev/vdb: data

  3. If there is no file system, create a file system using the mkfs -t command.

    mkfs -t command (without file system)
    ubuntu@volume-attach-1:~$ sudo mkfs -t xfs /dev/vdb
    meta-data=/dev/vdb isize=512 agcount=4, agsize=65536 blks
                = sectsz=512 attr=2, projid32bit=1
                = crc=1 finobt=1, sparse=1, rmapbt=0
                =reflink=1
    data = bsize=4096 blocks=262144, imaxpct=25
                = sunit=0 swidth=0 blks
    naming =version 2 bsize=4096 ascii-ci=0, ftype=1
    log =internal log bsize=4096 blocks=2560, version=2
                = sectsz=512 sunit=0 blks, lazy-count=1
    realtime =none extsz=4096 blocks=0, rtextents=0

  4. Confirm that the file system was created properly.

    file -s command (check file system creation)
    ubuntu@volume-attach-1:~$ sudo file -s /dev/vdb
    /dev/vdb: SGI XFS filesystem data (blksz 4096, inosz 512, v2 dirs)

  5. Use the mkdir command to create a mount point directory to use the volume.

    • The mount point is the location where files are read and written after the volume is mounted.
    mkdir command (Create mount point directory
    ubuntu@volume-attach-1:~$ sudo mkdir -p /data/vdb

  6. Use the mount command to attach the directory and volume created in the previous step.

    mount command (attaching directories and volumes)
    ubuntu@volume-attach-1:~$ sudo mount /dev/vdb /data/vdb

  7. Use the lsblk command to check whether the file system and mount point have been applied properly.

    lsblk command (check file system and mount point enforcement)
    ubuntu@volume-attach-1:~$ sudo lsblk
    NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
    vda 252:0 0 20G 0 disk /boot
    vdb 252:16 0 8G 0 disk /data/vdb

  8. To maintain the connected volume even after system reboot, first check the UUID of the device.

    lsblk command (check device's UUID)
    ubuntu@volume-attach-1:~$ lsblk -o +UUID
    NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT UUID
    vda 252:0 0 20G 0 disk /boot 876da551-d737-4350-9cd9-bb3fc9069bff
    vdb 252:16 0 1G 0 disk /data/vdb 2d332f0b-1771-4b83-bd5c-7e9d1cc91ed2

  9. Add the above content to the /etc/fstab file using the confirmed UUID.

    Keep volumes connected after system reboot
    UUID=2d332f0b-1771-4b83-bd5c-7e9d1cc91ed2 /data/vdb xfs defaults,nofail 0 2

Attach volume on Linux(xfs)​

For instances created with OS images (CentOS Stream 8, Rocky Linux 8, etc.) that use the Connection failure may occur. To use the volume normally, proceed with the following steps.

  1. First mount the file system using the nouuid option.

    • In case of temporary use, this problem can only be solved with this procedure.
    mount command (mount and nouuid options)
    [centos@centos-stream-8 test]$ sudo mount -t xfs -o nouuid /dev/vdb1 /home/centos/test

  2. Unmount.

    umount command (unmount)
    [centos@centos-stream-8 test]$ sudo umount /dev/vdb1

  3. Regenerate the uuid.

    • For a permanent solution, you must proceed with the relevant procedure.
    xfs_admin command (uuid regeneration)
    [centos@centos-stream-8 test]$ sudo xfs_admin -U generate /dev/vdb1
    Clearing log and setting UUID
    writing all SBs
    new UUID = ff714d9b-8c90-40e2-9e2f-12623220e1cf

  4. Set the newly created UUID to the restored snapshot volume.

    xfs_admin command (UUID update)
    [centos@centos-stream-8 test]$ sudo xfs_admin -U ff714d9b-8c90-40e2-9e2f-12623220e1cf /dev/vdb1
    Clearing log and setting UUID
    writing all SBs
    new UUID = ff714d9b-8c90-40e2-9e2f-12623220e1cf

  5. Mount the restored snapshot to the instance mount point.

    mount command (remount)
    [centos@centos-stream-8 test]$ sudo mount /dev/vdb1 /home/centos/test

Attach volume on Windows​

In a Windows environment, connect to the instance via remote desktop to initialize disks and allocate volumes.

  1. Start > Run or type the WIN + R shortcut keys.

  2. In the Run screen, enter diskmgmt.msc and click the [OK] button or press Enter.

  3. Select newly added disk in Disk Management, then right-click and select Online to bring it online.

  4. Select the newly added disk, right-click, and select Initialize Disk. When selected, a disk initialization pop-up window will appear.

  5. After checking the newly added disk, select the GPT partition table and click the [OK] button.

    • To connect a disk exceeding 2TB, you must select a GPT partition table.

  6. After selecting the initialized disk, right-click and select New Simple Volume. When selected, the Simple Volume Creation Wizard pops up.

  7. Run the Create Simple Volume Wizard.

  8. In the Simple Volume Creation Wizard, follow the instructions to select volume size, drive letter assignment, format options, etc., then exit to complete the volume connection.

Extend File system (Linux)​

In order to expand the size of a volume and store data in the added disk space, directly connect to the instance and update the size of the file system and partition. After the user expands the size of the volume, the process of using file system and partition related commands on the instance to expand the file system size is as follows. However, the required procedures may vary depending on the file system and partition configuration of your instance.

OS-specific partition-table format and file-system-type​
OS nameversionRoot volume partition table formatfile system
Ubuntu20.04GPT              ext4
Ubuntu22.04GPText4
Ubuntu24.04GPText4
CentOS Stream9MBRxfs
Rocky Linux8.10MBRxfs
Rocky Linux9.4MBRxfs
Alma Linux8.10GPTxfs
Alma Linux9.4GPTxfs
Windows2019MBRntfs
Windows2022MBRntfs
info

Changing the file system or partition does not modify or delete data on the volume. However, before changing file systems and partitions, we recommend backing up volumes containing important data so that you can revert the changes if necessary.

The system extension procedures applicable to the most widely used Linux file systems, ext4 and xfs, are as follows:

  • In the following Linux example, if there is an instance with a root volume of 50 GB that the user previously created, it explains how to expand the volume to increase the root volume of the instance to 80 GB and then expand the partition and file system.

Connect instance​

You can connect to your instance using the SSH command in the terminal. For details of the connection process, please refer to Connect to Linux instance.

Check partition and file system info​

You can use the lsblk and df commands to check the partition to be extended and file system information.

  • e.g. If you look at the partition information before volume expansion, you can see that the 50G-sized root volume is connected to the device name /dev/vda, and the 30G-sized non-root volume is connected to /dev/vdb. can. The root volume, /dev/vda device, has a vda1 partition of 50G size.

    lsblk command (before volume expansion)
    ubuntu@volume-test:~$ sudo lsblk
    NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
    vda 252:0 0 50G 0 disk
    ├─vda1 252:1 0 49.9G 0 part /
    ├─vda14 252:14 0 4M 0 part
    └─vda15 252:15 0 106M 0 part /boot/efi
    vdb 252:16 0 30G 0 disk /data

After expanding the size of the root volume, run the lsblk command to check the expanded root volume.

  • e.g. When executing the lsblk command after expanding the size of the root volume to 80G, the size of the /dev/vda device is expanded to 80G, but the size of the vda1 partition remains unchanged at 50G. You can check that.

    lsblk command (after expanding volume)
    ubuntu@volume-test:~$ sudo lsblk
    NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
    vda 252:0 0 80G 0 disk
    ├─vda1 252:1 0 49.9G 0 part /
    ├─vda14 252:14 0 4M 0 part
    └─vda15 252:15 0 106M 0 part /boot/efi
    vdb 252:16 0 30G 0 disk /data

You can check the file system type of the volume connected to the instance by running the df command.

  • e.g. The size of the /dev/vda1 partition is still 50G, and you can see information that it is using ext4.

    df command
    ubuntu@volume-test:~$ sudo df -Th
    Filesystem Type Size Used Avail Use% Mounted on
    ...
    /dev/vda1 ext4 49G 1.5G 47G 3% /
    ...
    /dev/vdb ext4 30G 45M 28G 1% /data

Extend partition​

You can use the growpart command to expand the size of a partition to its maximum size.

  • e.g. Run the growpart command to expand the size of the vda1 partition.

    Extend partition size
    ubuntu@volume-test:~$ sudo growpart /dev/vda 1

After extending, run the lsblk command to check the size of the extended partition.

  • e.g. After expansion, run the lsblk command to confirm that the size of the vda1 partition has been expanded to 80G.

    Confirm after extending partition
    ubuntu@volume-test:~$ lsblk
    NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
    vda 252:0 0 80G 0 disk
    ├─vda1 252:1 0 79.9G 0 part /

Extend file system​

You can expand the file system to store data in the expanded space.
Commands vary depending on the file system type. If you are using ex4, use the resize2fs command, and if you are using xfs, use the xfs_growfs command.

File system extensions (ext4)
sudo resize2fs /dev/vda1
File System Extensions (xfs)
sudo xfs_growfs -d /

Check the expanded file system with the df command. After you complete file system expansion, you can save data to the added disk space without having to reboot your instance.

  • Execute the df command to confirm that the /dev/vda1 file system has been expanded to 80G.

    df command (after file system expansion)*
    ubuntu@volume-test:~$ df -Th
    Filesystem Type Size Used Avail Use% Mounted on
    /dev/vda1 ext4 78G 1.6G 76G 2% /
    ...
    /dev/vdb ext4 30G 45M 28G 1% /data
    ...

Extend file system (Windows)​

Here's how to extend the file system in a Windows environment:

  • The following example assumes that there is an instance with a root volume of 50 GB that the user previously created, and explains how to increase the root volume of the instance to 70 GB using the volume expansion function and then expand the file system.

Expand file system with disk-management-tool​

Here's how to expand the file system through the Disk Management tool in a Windows environment:

  1. Go to KakaoCloud Console > Beyond Compute Service > GPU.

  2. In the Instance menu, select the instance's Show [More] icon > Use RDP to connect.

  3. Confirm your password and log in to your Windows instance.

  4. Right-click the [Windows] icon and select Disk Management.

  5. In Disk Management Tool, select Action > Refresh.

  6. Right-click the formatted partition on the disk you want to extend its size, and select Expand volume.

    • The partition must be adjacent to the right of the partition you want to extend, if it is not adjacent to the drive, Expand volume may be disabled. Additionally, volumes use the Master Boot Record (MBR) partition format and cannot exceed 2 TB.

  7. Click the [Next] button in the Expand volume Wizard.

  8. Enter the megabytes to expand in Select Space (MB) and click the [Next] button.

    • Typically specifies the maximum amount of space available.

  9. Click the [Finish] button and check the expanded file system.

Extend partition with PowerShell​

Here's how to extend a file system via PowerShell in a Windows environment:

  1. Go to KakaoCloud Console > Beyond Compute Service > GPU.

  2. In the Instance menu, select the instance's [More] icon > Use RDP to connect.

  3. Confirm your password and log in to your Windows instance.

  4. Right-click the [Windows] icon and select Windows PowerShell (Administrator).

  5. Check the drive letter of the partition to be extended.

    drive letter
    Get-Partition

  6. Rescan the disk.

    Disk Rescan
    "rescan" | diskpart

  7. Use the drive letter to determine the minimum and maximum allowed size (in bytes) of the partition.

    Disk Rescan
    Get-PartitionSupportedSize -DriveLetter <drive-letter>

  8. Extend the partition to the maximum available size.

    Partition maximum extension command
    Resize-Partition -DriveLetter <drive-letter> -Size $(Get-PartitionSupportedSize -DriveLetter <drive-letter>).SizeMax
    • To extend a partition to a specified size, enter the size in GB format in <size>.
    Partition specification size extension command
    Resize-Partition -DriveLetter <drive-letter> -Size <size>