You can resize a system disk or a data disk to extend its capacity. If your Linux Elastic Compute Service (ECS) instance does not support online resizing of disks, you can resize the disks on the instance offline. After you resize a disk on an instance offline, you must restart the instance for the changes to take effect. The restart operation interrupts services that run on the instance for a short period of time. We recommend that you resize disks offline during off-peak hours.

Prerequisites

The requirements described in the following table are met before you can resize a disk offline for a Linux instance.
Resource Requirement
Instance
Disk
  • The disk is in the In Use (In Use) state.
  • The disk is an enhanced SSD (ESSD), a standard SSD, or an ultra disk.
  • After you renew a subscription instance and downgrade its configurations, you cannot resize the subscription disks on the instance for the remainder of the current billing cycle.
Notice A disk in the master boot record (MBR) partition format cannot be resized to 2 TiB or larger. To resize an MBR disk to larger than 2 TiB, we recommend that you create a disk larger than 2 TiB, partition and format the new disk to the GUID Partition Table (GPT) format, and then copy data from the original MBR disk to the new GPT disk. For more information about how to partition and format disks to GPT, see Partition and format a data disk larger than 2 TiB in size.

Background information

The following table describes the resources used in the examples of this topic.
Resource Description
Image used by the instance Alibaba Cloud Linux 2.1903 LTS 64-bit public image
System disk /dev/vda: uses the MBR partition format and the ext4 file system, and is resized from 40 GiB to 60 GiB.
Data disks
  • /dev/vdb: uses the MBR partition format and the ext4 file system, and is resized from 40 GiB to 60 GiB.
  • /dev/vdc: uses the GPT partition format and the xfs file system, and is resized from 40 GiB to 60 GiB.

Step 1: Create a snapshot

Create a snapshot for a disk to back up disk data before you resize the disk.

  1. Log on to the ECS console.
  2. In the left-side navigation pane, choose Instances & Images > Instances.
  3. In the top navigation bar, select a region.
  4. Find the instance whose disk you want to resize and click the instance ID.
  5. On the Instance Details page, click the Cloud Disk tab.
  6. Find the disk that you want to resize and click Create Snapshot in the Actions column.
  7. In the Create Snapshot dialog box, enter a snapshot name, specify tags, and then click Create.
  8. Click the Snapshot tab to view the snapshot.
    When the snapshot is created, 100% appears in the Progress column corresponding to the snapshot. You can proceed to perform subsequent operations.

Step 2: Resize the disk and restart or start the instance in the ECS console

  1. On the Instance Details page, click the Cloud Disk tab.
  2. Find the disk that you want to resize and choose More > Resize Disk in the Actions column.

    To batch resize disks, log on to the ECS console by using your Alibaba Cloud account and choose Storage & Snapshots > Disks in the left-side navigation pane. On the Disks page, select the disks that you want to resize and click Resize Disk in the lower part of the page. Disks that are attached to the same instance cannot be resized together.

  3. On the Resize Disks page, set the Size after Resize parameter.
    The specified Size after Resize value must be greater than the current capacity.
  4. Verify the price. Read and select ECS Service Terms, and then click Confirm.
  5. Read the notes, click I have read the notes. Resize, and then complete the payment.
  6. Restart or start the instance in the ECS console.
    Note
    • If your instance is in the Running (Running) state, you must restart the instance in the ECS console. If your instance is in the Stopped (Stopped) state, you must start the instance in the ECS console.
    • The disk resizing operation takes effect only after you restart or start the instance by using the console or by calling the RebootInstance operation. If you restart the instance from within the operating system, the disk resizing operation does not take effect.

    The following procedure describes how to restart an instance.

    1. In the left-side navigation pane, choose Instances & Images > Instances.
    2. On the Instances page, find the instance that you want to restart and choose More > Instance Status > Restart in the Actions column.
    3. In the Restart Instance dialog box, click OK.

Step 3: View the disk partitions

Log on to the instance to view the partition types (MBR and GPT) and file system types (such as ext4 and xfs) of the system disk and data disks. Subsequent resizing operations vary based on the types of the partitions and file systems.

  1. Log on to the instance. For more information, see Connect to a Linux instance by using password authentication.
  2. Run the following command to view the disks attached to the instance:
    fdisk -lu
    The following command output shows a partition of the system disk (/dev/vda1) and two partitions of data disks (/dev/vdb1 and /dev/vdc1): View the partitions of the disks
    No. Partition Description
    /dev/vda1 A partition of the system disk. A value of Linux for System indicates that the partition is in the MBR format.
    /dev/vdb1 A partition of a data disk. A value of Linux for System indicates that the partition is in the MBR format.
    /dev/vdc1 A partition of a data disk. A value of GPT for System indicates that the partition is in the GPT format.
    Note If the capacity of the system disk (/dev/vda) in the command output is still 40 GiB (Disk /dev/vda: 42.9 GB), the disk was not resized. We recommend that you restart the instance in the ECS console.
  3. Run the following command to check the file system types of existing partitions:
    df -Th

    A command output similar to the following one is displayed. In the command output, the Mounted on column lists the mount points corresponding to the file systems.

    View the file systems

Step 4: Resize partitions

When you view the disk partitions, you may find that the partitions and file systems within the instance are not resized. This step describes how to resize the partitions of the resized disk within the instance.

  1. Install the gdisk tool on the instance.
    You must perform this step if the partitions are in the GPT format. If the partitions are in the MBR format, skip this step.
    yum install gdisk -y
  2. Install the growpart tool.
    • If the instance runs a CentOS 7 or later operating system, run the following command:
      yum install -y cloud-utils-growpart
    • If the instance runs a Debian 8 or later operating system or an Ubuntu 14 or later operating system, run the following command:

      Update the software repository.

      apt-get update

      Install cloud-guest-utils.

      apt-get install -y cloud-guest-utils
  3. Run the following command to resize partitions:
    growpart /dev/vda 1
    In this example, the partition of the system disk is resized. Use a space to separate /dev/vda from 1. To resize other partitions, modify the command accordingly. A command output similar to the following one is displayed: growpart
    Note When you run the growpart /dev/vda 1 command, the unexpected output in sfdisk --version [sfdisk, from util-linux 2.23.2] error may appear. For information about how to troubleshoot this problem, see the "FAQ" section of this topic.

Step 5: Resize file systems

This step describes how to resize the file systems of partitions within the instance.

  1. Resize the file systems within the instance based on the file system types that you obtained.
    • To resize ext* (such as ext4) file systems, run the following commands.
      To resize the file system of the /dev/vda1 partition of the system disk, run the following command:
      resize2fs /dev/vda1    
      To resize the file system of the /dev/vdb1 partition of a data disk, run the following command:
      resize2fs /dev/vdb1          
      Note /dev/vda1 and /dev/vdb1 are partition names. Replace them with your actual partition names.
    • To resize the xfs file system of the /dev/vdc1 partition of a data disk, run the following command:
      xfs_growfs /media/vdc
      Note /media/vdc is the mount point of the /dev/vdc1 partition. Replace it with the actual mount point of your partition.
    • To resize the btrfs file system of a data disk, run the following command:
      btrfs filesystem resize max /mountpoint
      Note /mountpoint is the mount point of the data disk. Replace it with the actual mount point of your data disk.
  2. Run the following command to check the resize results:
    df -Th

    A command output similar to the following one is displayed:

    Check the resize results
    After you resize the file systems, check whether their actual sizes are consistent with the new specified values.
    • If the file system sizes are increased to the new specified values and the business programs on the instance can run normally, the resize operation is complete.
    • If the file system sizes are not increased to the new specified values, use the snapshot that you created in Step 1 to roll back the disk.

FAQ

  • Question: When the growpart /dev/vda 1 command is run on a Linux instance, the unexpected output in sfdisk --version [sfdisk, from util-linux 2.23.2] error appears. What do I do?
    Answer:
    1. Run the locale command to view the character encoding type of the instance. If the character encoding type is not en_US.UTF-8, switch it to en_US.UTF-8.
      1. Run the following command to switch the character encoding type:
        LANG=en_US.UTF-8
      2. If the problem persists, run the following command to switch the character encoding type:
        export LC_ALL=en_US.UTF-8
      3. If the problem still persists, run the following command to switch the character encoding type:
        localectl set-locale LANG=en_US.UTF-8
      4. If the preceding solutions do not solve the problem, run the following command to switch the character encoding type:
        export LANGUAGE=en_US.UTF-8
    2. If the problem remains unsolved, run the reboot command to restart the instance.
    Notice

    If the partition is resized after you switch the character encoding type, we recommend that you switch back to the original character encoding type.

  • Question: When the growpart /dev/vda 1 command is run on the Linux instance, the -bash: growpart: command not found error appears. What do I do?
    Answer:
    1. Run the uname -a command to check the version of the Linux kernel. The procedure described in this topic applies to Linux kernel 3.6.0 and later.

      If the version of the Linux kernel is earlier than 3.6.0, resize the partitions of disks after you resize the disks. For more information, see Procedure for instances with kernels earlier than 3.6.0 and Resize partitions and file systems of Linux data disks.

    2. Install the growpart tool.
      • If the instance runs a CentOS 7 or later operating system, run the following command:
        yum install -y cloud-utils-growpart
      • If the instance runs a Debian 8 or later operating system or an Ubuntu 14 or later operating system, run the following command:
        apt install -y cloud-guest-utils
  • Question: The growpart tool used to resize partitions cannot be installed on CentOS 6.5. Why?
    Answer: The Linux kernel version of CentOS 6 is earlier than 3.6.0. To install the growpart tool on CentOS 6, perform the following steps:
    1. Change the source address of the YUM repository on CentOS 6. For more information, see Change the CentOS 6 source address.
      Note CentOS 6 has reached its end of life (EOL). To install the software package on CentOS 6 by using YUM, change the source address of the YUM repository.
    2. Install the dracut-modules-growroot tool on instances that run CentOS 6 to resize the partitions of disks. For more information, see Procedure for instances with kernels earlier than 3.6.0.

Other scenarios for disk resizing