Help Center/ Elastic Cloud Server/ Getting Started/
Updated on 2024-09-06 GMT+08:00
View PDF
Share

Initializing a Linux Data Disk Larger Than 2 TiB (parted)

Scenarios

This section uses CentOS 7.4 64bit to describe how to use parted to initialize a data disk whose capacity is larger than 2 TiB. In the following operations, the capacity of the example disk is 3 TiB.

The maximum partition size that MBR supports is 2 TiB and that GPT supports is 18 EiB. If the disk size you need to partition is greater than 2 TiB, partition the disk using GPT.

The fdisk partitioning tool is suitable only for MBR partitions, and the parted partitioning tool is suitable for both MBR and GPT partitions. For more information, see Scenarios and Disk Partitions.

The method for initializing a disk varies slightly depending on the OS running on the server. This document is used for reference only. For the detailed operations and differences, see the product documents of the corresponding OS.

When using a disk for the first time, if you have not initialized it, including creating partitions and file systems, the additional space added to this disk in an expansion later may not be normally used.

Prerequisites

  • A data disk has been attached to a server and has not been initialized.
  • You have logged in to the server.

Creating and Mounting a Partition

The following example shows you how a new partition can be created on a new data disk that has been attached to a server. The partition will be created using parted, and GPT will be used. Furthermore, the partition will be formatted using the ext4 file system, mounted on /mnt/sdc, and configured to mount automatically at startup.

  1. Query information about the new data disk.

    lsblk

    Information similar to the following is displayed:

    [root@ecs-centos74 ~]# lsblk
NAME   MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
vda    253:0    0  40G  0 disk
├─vda1 253:1    0   1G  0 part /boot
└─vda2 253:2    0  39G  0 part /
vdb    253:16   0   3T  0 disk

    In the command output, this server contains two disks. /dev/vda and /dev/vdb. /dev/vda is the system disk, and /dev/vdb is the new data disk.

  2. Launch parted to partition the new data disk.

    parted New data disk

    In this example, run the following command:

    parted /dev/vdb

    Information similar to the following is displayed:

    [root@ecs-centos74 ~]# parted /dev/vdb
GNU Parted 3.1
Using /dev/vdb
Welcome to GNU Parted! Type 'help' to view a list of commands.
(parted)

  3. Enter p and press Enter to view the current disk partition style.

    Information similar to the following is displayed:

    (parted) p
Error: /dev/vdb: unrecognised disk label
Model: Virtio Block Device (virtblk)
Disk /dev/vdb: 3299GiB
Sector size (logical/physical): 512B/512B
Partition Table: unknown
Disk Flags:
(parted)

    In the command output, the Partition Table value is unknown, indicating that no partition style is set for the new disk.

  4. Set the disk partition style.

    mklabel Disk partition style

    The disk partition style can be MBR or GPT. If the disk capacity is greater than 2 TiB, use GPT.

    mklabel gpt

    The maximum disk size supported by MBR is 2 TiB, and that supported by GPT is 18 EiB. Because an EVS data disk currently supports up to 32 TiB, use GPT if your disk size is greater than 2 TiB.

    If the partition style is changed after the disk has been used, all data on the disk will be lost, so take care to select an appropriate partition style when initializing the disk. If you must change the partition style to GPT after a disk has been used, it is recommended that you back up the disk data before the change.

  5. Enter p and press Enter to view the disk partition style.

    Information similar to the following is displayed:

    (parted) mklabel gpt
(parted) p
Model: Virtio Block Device (virtblk)
Disk /dev/vdb: 3299GiB
Sector size (logical/physical): 512B/512B
Partition Table: gpt
Disk Flags:
 
Number  Start  End  Size  File system  Name  Flags
 
(parted)

  6. Enter unit s and press Enter to set the measurement unit of the disk to sector.
  7. Create a new partition.

    mkpart Partition name Start sector End sector

    In this example, run the following command:

    mkpart opt 2048s 100%

    In this example, one partition is created for the new data disk, starting on 2048 and using 100% of the rest of the disk. The two values are used for reference only. You can determine the number of partitions and the partition size based on your service requirements.

    Information similar to the following is displayed: 
    (parted) mkpart opt 2048s 100%
Warning: The resulting partition is not properly aligned for best performance.
Ignore/Cancel? Ignore

    If the preceding warning message is displayed, enter Ignore to ignore the performance warning.

  8. Enter p and press Enter to print the partition details.

    Information similar to the following is displayed:

    (parted) p
Model: Virtio Block Device (virtblk)
Disk /dev/vdb: 6442450944s
Sector size (logical/physical): 512B/512B
Partition Table: gpt
Disk Flags:
 
Number  Start  End          Size         File system  Name  Flags
 1      2048s  6442448895s  6442446848s               opt

    Details about the dev/vdb1 partition are displayed.

  9. Enter q and press Enter to exit parted.
  10. View the disk partition information.

    lsblk

    Information similar to the following is displayed:

    [root@ecs-centos74 ~]# lsblk
NAME   MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
vda    253:0    0  40G  0 disk
├─vda1 253:1    0   1G  0 part /boot
└─vda2 253:2    0  39G  0 part /
vdb    253:16   0   3T  0 disk
└─vdb1 253:17   0   3T  0 part

    In the command output, /dev/vdb1 is the partition you created.

  11. Format the new partition with a desired file system format.

    mkfs -t File system format /dev/vdb1

    In this example, the ext4 format is used for the new partition.

    mkfs -t ext4 /dev/vdb1

    Information similar to the following is displayed:

    [root@ecs-centos74 ~]# mkfs -t ext4 /dev/vdb1
mke2fs 1.42.9 (28-Dec-2013)
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
Stride=0 blocks, Stripe width=0 blocks
201326592 inodes, 805305856 blocks
40265292 blocks (5.00%) reserved for the super user
First data block=0
Maximum filesystem blocks=2952790016
24576 block groups
32768 blocks per group, 32768 fragments per group
8192 inodes per group
Superblock backups stored on blocks:
        32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208,
        4096000, 7962624, 11239424, 20480000, 23887872, 71663616, 78675968,
        102400000, 214990848, 512000000, 550731776, 644972544
 
Allocating group tables: done
Writing inode tables: done
Creating journal (32768 blocks): done
Writing superblocks and filesystem accounting information: done

    The formatting takes a period of time. Observe the system running status and do not exit.

    The partition sizes supported by file systems vary. Choose an appropriate file system format based on your service requirements.

  12. Create a mount point.

    mkdir Mount point

    In this example, the /mnt/sdc mount point is created.

    mkdir /mnt/sdc

    The /mnt directory exists on all Linux systems. If the mount point cannot be created, it may be that the /mnt directory has been accidentally deleted. You can run mkdir -p /mnt/sdc to create the mount point.

  13. Mount the new partition on the created mount point.

    mount Disk partition Mount point

    In this example, the /dev/vdb1 partition is mounted on /mnt/sdc.

    mount /dev/vdb1 /mnt/sdc

  14. Check the mount result.

    df -TH

    Information similar to the following is displayed:

    [root@ecs-centos74 ~]# df -TH
Filesystem     Type      Size  Used Avail Use% Mounted on
/dev/vda2      ext4       42G  1.5G   38G   4% /
devtmpfs       devtmpfs  2.0G     0  2.0G   0% /dev
tmpfs          tmpfs     2.0G     0  2.0G   0% /dev/shm
tmpfs          tmpfs     2.0G  8.9M  2.0G   1% /run
tmpfs          tmpfs     2.0G     0  2.0G   0% /sys/fs/cgroup
/dev/vda1      ext4      1.1G  153M  801M  17% /boot
tmpfs          tmpfs     398M     0  398M   0% /run/user/0
/dev/vdb1      ext4      3.3T   93M  3.1T   1% /mnt/sdc

    You should now see that partition /dev/vdb1 is mounted on /mnt/sdc.

Configuring Automatic Mounting at System Start

The fstab file controls what disks are automatically mounted at server startup. You can configure the fstab file of a server that has data. This operation will not affect the existing data.

The following example uses UUIDs to identify disks in the fstab file. You are advised not to use device names (like /dev/vdb1) to identify disks in the file because device names are assigned dynamically and may change (for example, from /dev/vdb1 to /dev/vdb2) after a server stop or start. This can even prevent your server from booting up.

UUIDs are the unique character strings for identifying partitions in Linux.

  1. Query the partition UUID.

    blkid Disk partition

    In this example, the UUID of the /dev/vdb1 partition is queried.

    blkid /dev/vdb1

    Information similar to the following is displayed:

    [root@ecs-test-0001 ~]# blkid /dev/vdb1
/dev/vdb1: UUID="0b3040e2-1367-4abb-841d-ddb0b92693df" TYPE="ext4"

    Carefully record the UUID, as you will need it for the following step.

  2. Open the fstab file using the vi editor.

    vi /etc/fstab

  3. Press i to enter editing mode.
  4. Move the cursor to the end of the file and press Enter. Then, add the following information: 

    UUID=0b3040e2-1367-4abb-841d-ddb0b92693df /mnt/sdc                ext4    defaults        0 2
    The preceding information is used for reference only. The line starting with UUID is the information added. Edit this line from left to right to match the following format:
    • UUID: The UUID obtained in 1.
    • Mount point: The directory on which the partition is mounted. You can query the mount point using df -TH.
    • Filesystem: The file system format of the partition. You can query the file system format using df -TH.
    • Mount option: The partition mount option. Usually, this parameter is set to defaults.
    • Dump: The Linux dump backup option.
      • 0: Linux dump backup is not used. Usually, dump backup is not used, and you can set this parameter to 0.
      • 1: Linux dump backup is used.
    • fsck: The fsck option, which means whether to use fsck to check the disk during startup.
      • 0: not use fsck.
      • If the mount point is the root partition (/), this parameter must be set to 1.

        If this parameter is set to 1 for the root partition, this parameter for other partitions must start with 2 because the system checks the partitions in the ascending order of the values.

  5. Press Esc, enter :wq, and press Enter.

    The system saves the configurations and exits the vi editor.

  6. Verify that the disk is auto-mounted at startup.

    1. Unmount the partition.

      umount Disk partition

      In this example, run the following command:

      umount /dev/vdb1

    2. Reload all the content in the /etc/fstab file.

      mount -a

    3. Query the file system mounting information.

      mount | grep Mount point

      In this example, run the following command:

      mount | grep /mnt/sdc

      If information similar to the following is displayed, automatic mounting has been configured:

      root@ecs-test-0001 ~]# mount | grep /mnt/sdc
/dev/vdb1 on /mnt/sdc type ext4 (rw,relatime,data=ordered)