Help Center/ Elastic Volume Service/ User Guide/ Purchasing and Using an EVS Disk/ Initializing EVS Data Disks/ Initialization Overview
Updated on 2024-08-08 GMT+08:00
View PDF

Initialization Overview

After you attach a new data disk to a server, you must initialize the disk including creating partitions, creating file systems, and mounting the partitions before you can use the disk.

Scenarios

  • System disk

    When a server is created, a system disk is automatically initialized with master boot record (MBR).

  • New data disk
    • If a data disk is created together with a server, EVS automatically attaches it to the server. You only need to initialize it to make it available for use.
    • If a data disk is created explicitly, you need to first attach it to a server and then initialize it.

    For detailed operation instructions, see Table 1.

  • Existing data disk
    An existing data disk is a disk created from a snapshot, a backup, or an image, or a disk detached from another server.
    • To use the disk's existing partitions:
    • To re-initialize the disk:

      Re-partitioning a disk will erase all the existing data on the disk, so you are advised to use snapshots to back up the disk data first.

      • In Linux, unmount the partitions, delete them (by running fdisk Disk name, entering d and the partition number, and entering w), and then re-initialize the disk.
      • In Windows, delete the partitions (using the volume deletion tool) and then re-initialize the disk. For detailed initialization operations, see Table 1.

      Initializing a disk does not delete the snapshots created for the disk, so you can still use snapshots to roll back data to the source disk after the disk is initialized.

Prerequisites

  • You have attached the disk to a server.
  • You are advised to back up the disk data before re-initializing a disk, because data may be lost after a disk is re-initialized.
  • You have logged in to the server.

Notes and Constraints

  • A disk created from a data source does not need to be initialized. Such a disk contains the source data in the beginning. Initializing the disk may clear the initial data on it.
  • Initializing a disk does not delete the snapshots created for the disk, so you can still use snapshots to roll back data to the source disk after the disk is initialized.

Operation Instructions

Table 1 Disk initialization instructions

Disk Capacity

Partition Style

Partition Type

OS

Reference

Capacity ≤ 2 TiB

GPT or MBR
  • GPT partitions are not classified, and there is no limit on the number of GPT partitions.
  • MBR partitions can be:
    • Four primary partitions
    • Three primary partitions and one extended partition

      The number of logical partitions allowed in the extended partition is not limited, so theoretically you can create as many logical partitions as you want.

    If you need five or more partitions, use the "primary partitions + one extended partition" model and then create logical partitions in the extended partition.

Linux

Initializing a Linux Data Disk (Less Than or Equal to 2 TiB)

Windows

Initializing a Windows Data Disk

Capacity > 2 TiB

GPT

GPT partitions are not classified, and there is no limit on the number of GPT partitions.

Linux

Initializing a Linux Data Disk (Greater Than 2 TiB)

Windows

Initializing a Windows Data Disk
  • The maximum disk size that MBR supports is 2 TiB, and that GPT supports is 18 EiB. If your disk is greater than 2 TiB or you may expand it to over 2 TiB later, use GPT when initializing disks.
  • If you change the partition style of a disk, data on the disk will be erased. Select an appropriate partition style when initializing disks.
  • In Linux, you can use either fdisk or parted to create MBR partitions, and use only parted to create GPT partitions.
Parent Topic: Initializing EVS Data Disks