How to fix Error 3247 on Windows Server 2025 / Windows 11

Troubleshooting Guide: SQL Server Error 3247 – Wrong Media Sequence Number

Why This Error Happens

Error 3247 occurs when SQL Server 2025 is performing a restore operation from a multi-volume backup set and detects that the media volume currently loaded does not match the expected sequence number in the backup header. This typically happens when backup tapes, files, or disk volumes belonging to a striped or multi-file backup set are presented to the RESTORE command out of order. It can also occur if a backup file was manually renamed, partially copied, or corrupted, causing SQL Server’s media management subsystem to misread the volume’s internal sequence metadata.

Step-by-Step Resolution Guide

Step 1: Identify the Correct Volume Order

  • Run RESTORE LABELONLY FROM DISK = 'path_to_backup' or RESTORE HEADERONLY against each file in the backup set to inspect the MediaSequenceNumber and FamilySequenceNumber columns.
  • Record the correct numerical order of each .bak or .trn file as indicated by SQL Server metadata, not by filename or timestamp.
  • On Windows Server 2025/Windows 11, ensure all volume files are accessible from the same local path or UNC share (e.g., \\BackupServer\SQLBackups\).

Step 2: Reissue the RESTORE Command in Correct Sequence

  • Use the RESTORE DATABASE ... FROM DISK = N'Volume1', DISK = N'Volume2' syntax, explicitly listing files in the sequence number order identified in Step 1.
  • If using a media/tape library, physically reload the volumes in the exact order requested by the error message prompt (e.g., “insert volume 2”).
  • Example T-SQL:
RESTORE DATABASE [YourDB] 
FROM DISK = N'C:\Backups\Backup_Vol1.bak', 
     DISK = N'C:\Backups\Backup_Vol2.bak'
WITH FILE = 1, NORECOVERY;

Step 3: Validate Backup File Integrity

  • Run RESTORE VERIFYONLY FROM DISK = 'path_to_backup' on each volume to confirm the file is not corrupted or truncated.
  • Check file properties in Windows File Explorer or via PowerShell (Get-Item <path> | Select Length) to ensure no volume was partially copied during a file transfer.
  • If a volume is missing or corrupted, re-copy it from the original source media or backup repository.

Step 4: Rebuild the Backup Set Reference (If Using a Media Set)

  • If the backup was created with WITH MEDIANAME or as part of a named media set, confirm the MEDIANAME and MEDIAPASSWORD (if applicable) match across all volumes using RESTORE HEADERONLY.
  • Avoid mixing volumes from different backup operations into a single RESTORE statement, as this causes sequence mismatches even if filenames appear similar.

Alternative or Verification Steps

  • After a successful restore, verify database consistency by running DBCC CHECKDB('YourDB') WITH NO_INFOMSGS to confirm no corruption was introduced during the restore process.
  • If the correct sequence still fails, take a fresh full backup of the source database (if accessible) using BACKUP DATABASE ... TO DISK with a single destination file to eliminate multi-volume sequencing issues entirely.