What Are the Restrictions on Using MgC for Storage Migration?

Objects with multiple versions

By default, only the latest version of objects in source buckets is migrated.

Storage class of target buckets

The storage class of target buckets can only be Standard or Infrequent Access. You can change the storage class of target buckets after the migration is complete.

Migration object

• Object names cannot contain special characters.
• The size limit for a single object is 4.76837158203125 TB (500 MB × 10,000). If the limit is exceeded, the migration may fail.

Migration network

Migrations are supported over public networks, private networks, and private lines.

Symbolic links

• Symbolic links cannot be used for specifying migration paths which define the migration scope. To migrate a path pointed to by a symbolic link, you need to:
  • Enter the actual path when specifying the migration path.
  • After the migration is complete, manually create a symbolic link to the path at the target.
• For migration from NAS_SMB or migration from NAS_NFS to OBS, symbolic links cannot be migrated.
• For migration from NAS_NFS to NAS_NFS, symbolic links can be migrated by enabling metadata migration. Otherwise, these files will be skipped during the migration.
• For migration from Alibaba Cloud OSS to NAS_NFS, symbolic links can be migrated by enabling metadata migration. Otherwise, the symbolic links will lose their link functionality and become regular files after the migration.
  NOTICE:

  If the objects that symbolic links point to are not completely migrated to the target, these symbolic link files may fail the verification. As a result, the task will be in a failed status. In this case, wait until the involved objects are completely migrated to the target, and try the task again.

Hard links

For migration from NAS_NFS to NAS_NFS, hard links can be migrated by enabling metadata migration. Otherwise, these files will be skipped during the migration.

Migration scope

You can migrate a single bucket or multiple buckets in batches.

Metadata migration

• Only Chinese characters, English characters, digits, and hyphens (-) can be migrated. Other characters cannot be migrated.
  • Chinese characters are URL-encoded during the migration.
    CAUTION:

    Chinese punctuation marks cannot be URL-encoded during the migration. If metadata contains Chinese punctuation marks, the metadata and the corresponding object will fail to be migrated.

  • English characters, digits, and hyphens (-) are directly migrated without being encoded.
• For heterogeneous migrations, metadata cannot be migrated.

Archived data

To migrate archived data from object storage, you must restore it first. You need to:

• Ensure the restoration is complete before creating migration workflows.
• Configure a validity period for restored data based on the total amount of data to be migrated. This helps prevent migration failures because restored data becomes archived again during the migration.
• Pay your source cloud vendor for restoring archived data. To learn about the pricing details, contact your source cloud vendor.

Concurrent subtasks

User-defined. There cannot be more than 10 concurrent subtasks for each online migration node.

For example, if there are 2 online migration nodes, the maximum number of subtasks can be 20 or any number below.

Object list files

These files must be stored in the same region as the target bucket.

• These files must be in .txt format, and their metadata Content-Type must be text/plain. The directory storing these files can only contain .txt files.
• A single file can contain a maximum of 100,000 lines.
• A single file cannot exceed 300 MB.
• A maximum of 10,000 list files can be stored in the folder.
• The files must be in UTF-8 without BOM.
• The length of each line in a file cannot exceed 65,535 characters, or the migration will fail.
• The Content-Encoding metadata of the files must be left empty, or the migration will fail.
• In the files, a tab character (\t) must be used to separate the URL and new file name in each line. The format is [URL][Tab character][New file name at the target]. Only the Chinese and special characters in the source and target file names must be URL encoded.
• Spaces are not allowed in each line in a file. Spaces may cause migration failures because they may be mistakenly identified as object names.

Migration source: SMB systems

• File systems where a single directory contains more than 5 million files cannot be migrated.
• Resumable transfer is not supported.
• Symbolic links cannot be migrated.

Migration source: NAS file systems

• The following types of files can be migrated: regular files, directories, symbolic link files, and hard link files.
  CAUTION:

  If the file handle of a source file is occupied or the source file is deleted, the migration of the file will fail.

• Special files such as character device files, block device files, sockets, and pipe files cannot be migrated.
• The metadata of symbolic link files cannot be migrated.

Migration from non-SMB file systems to SMB file systems

In SMB file systems on Windows, file names are case-insensitive, so files with names that differ only in letter casing cannot coexist in the same directory. If such files exist at the source, you need to rename them before the migration. Otherwise, only the last uploaded files will be retained at the target.

For example, if a source directory contains two files dir/A.txt and dir/a.txt, and dir/A.txt is last uploaded to the target, file dir/a.txt will be retained at the target, but its content will be that of dir/A.txt.