A high-performance PowerShell script that recursively scans directories and calculates folder sizes, similar to TreeSize. Uses parallel processing with runspaces for efficient scanning of large disk drives.

• Fast parallel directory scanning using PowerShell runspaces
• Thread-safe CSV output with real-time results
• Detailed logging of access errors and skipped items
• Configurable depth limits and exclusion patterns
• Handles junction points and symbolic links
• Outputs sizes in Bytes, MB, and GB
• Progress tracking during scan operations

• PowerShell 7.0 or higher
• Windows 10/11 (tested on PowerShell 7.5)
• .NET Framework (included with PowerShell)

• OutputCsv (Mandatory): Path to the output CSV file where results will be written.

• Path: Directory path to scan. Default: Current directory (.)
• MaxDepth: Maximum recursion depth. Default: 0 (unlimited)
• Exclude: Array of wildcard patterns to exclude from scanning. Default: Empty array
• RecurseLinkBehavior: How to handle symbolic links and junction points. Options: Skip (default) or Follow
• Parallelism: Number of parallel worker threads. Default: ProcessorCount * 2
• VerboseLog: Path to the verbose log file. Default: C:\Temp\TreeSizeVerbose.log

Scan the current directory and output to a CSV file:

.\get-treesize.ps1 -OutputCsv ".\results.csv"

Scan the C: drive with 8 parallel workers:

.\get-treesize.ps1 -Path 'C:\' -Parallelism 8 -OutputCsv 'C:\Reports\c_drive_sizes.csv'

Scan a directory excluding certain folders:

.\get-treesize.ps1 -Path 'D:\Projects' `
    -Exclude @('*\node_modules\*', '*\.git\*', '*\bin\*') `
    -OutputCsv 'D:\project_sizes.csv'

Scan only 3 levels deep:

.\get-treesize.ps1 -Path 'C:\Users' -MaxDepth 3 -OutputCsv 'C:\users_tree.csv'

.\get-treesize.ps1 `
    -Path 'C:\' `
    -MaxDepth 0 `
    -Exclude @('*\Windows\*', '*\$Recycle.Bin\*') `
    -RecurseLinkBehavior 'Skip' `
    -Parallelism 8 `
    -OutputCsv 'C:\Reports\c_drive_sizes.csv' `
    -VerboseLog 'C:\Reports\treesize.log'

The CSV file contains the following columns:

• Path: Full path to the directory
• Depth: Recursion depth (0 = root level)
• SizeBytes: Total size in bytes
• SizeMB: Total size in megabytes (rounded to 2 decimals)
• SizeGB: Total size in gigabytes (rounded to 2 decimals)
• FileCount: Number of files in the directory (not including subdirectories)
• Timestamp: ISO 8601 timestamp of when the directory was scanned

Path,Depth,SizeBytes,SizeMB,SizeGB,FileCount,Timestamp
"C:\Users",0,1234567890,1177.37,1.15,1250,2024-01-15T10:30:45.1234567Z
"C:\Users\John",1,987654321,941.89,0.92,850,2024-01-15T10:30:46.2345678Z

• Uses concurrent collections (ConcurrentDictionary, ConcurrentQueue) for thread-safe operations
• Implements reader-writer locks for safe file writing
• Default parallelism is set to ProcessorCount * 2 for optimal performance
• Progress is displayed in real-time during scanning

• Access denied errors are logged but do not stop the scan
• Inaccessible files and directories are skipped
• All errors are written to the verbose log file with timestamps
• The script continues processing even if individual directories fail

The verbose log file contains:

• Timestamped entries for each operation
• Excluded directories
• Skipped reparse points (junction points/symbolic links)
• File enumeration errors
• Directory access errors

• The script creates output directories automatically if they don't exist
• Large drives may take considerable time to scan completely
• Use exclusion patterns to skip known large directories (e.g., node_modules, .git)
• Junction points and symbolic links are skipped by default to prevent infinite loops

Version 2.2 - Tested on PowerShell 7.5 / Windows 10-11

This script is provided as-is for use in scanning and analyzing disk usage.