How to fix Error 3445 on Windows Server 2025 / Windows 11

by

Troubleshooting Guide: Error 3445 – Invalid Undo File in SQL Server 2025

Why This Error Happens

Error 3445 occurs when SQL Server attempts to apply a transaction log backup using STANDBY mode, but the specified undo (.tuf) file is either corrupted, missing, or does not match the database ID it was originally created for. This typically happens when the undo file path has changed, the file was manually deleted or overwritten, or a mismatched undo file from a different database/restore sequence is being referenced. File system permission issues on the undo file directory can also prevent SQL Server from reading the file’s header correctly, triggering this validation failure.

Step-by-Step Resolution Guide

Step 1: Verify the Undo File Path and Existence

  • Confirm the exact path specified in the WITH STANDBY = 'path\filename.tuf' clause of your restore command.
  • Navigate to the folder (e.g., C:\SQLUndo\ or D:\MSSQL\Undo\) using File Explorer or PowerShell to confirm the .tuf file physically exists.
  • Run Get-ChildItem -Path "C:\SQLUndo\" -Filter *.tuf in PowerShell to list all undo files and check timestamps.

Step 2: Check File Permissions and Service Account Access

  • Ensure the SQL Server service account (visible in SQL Server Configuration Manager) has Read/Write/Modify permissions on the undo file directory.
  • Right-click the folder, go to Properties > Security, and add the SQL Server service account (e.g., NT Service\MSSQLSERVER) with Full Control if missing.
  • Restart the SQL Server service after updating permissions: Restart-Service -Name 'MSSQLSERVER'.

Step 3: Discard the Corrupted Undo File and Restart the Restore Sequence

  • Stop the restore sequence and run RESTORE DATABASE [YourDB] WITH RECOVERY to roll back and clear the invalid standby state if the database is stuck in restoring mode.
  • Delete the corrupted or mismatched .tuf file manually from the directory.
  • Re-run the log restore command with a freshly specified undo file path, ensuring it is unique per database: RESTORE LOG [YourDB] FROM DISK = 'logbackup.trn' WITH STANDBY = 'C:\SQLUndo\YourDB_new.tuf';

Step 4: Validate Database ID Consistency

  • Query SELECT name, database_id FROM sys.databases; to confirm the database ID matches what the undo file expects.
  • If database ID mismatches occur (common after restoring to a different server), avoid reusing undo files across different database instances or IDs.

Alternative or Verification Steps

  • Verify the fix by running the restore command again and confirming the database enters STANDBY mode without errors; check sys.databases for state_desc = 'STANDBY'.
  • If the issue persists, switch to WITH NORECOVERY instead of STANDBY for the log restore chain, then apply WITH RECOVERY only on the final log file to bypass undo file dependency entirely.