Lesson Learned #446: Simplifying SQLPackage Log Debugging with PowerShell
Published Oct 20 2023 10:00 AM 1,649 Views

Abstract:

Handling massive SQLPackage diagnostic logs, like those spanning over 4 million rows, can be an overwhelming task when troubleshooting support cases. This article introduces a PowerShell script designed to efficiently parse through SQLPackage diagnostic logs, extract error messages, and save them to a separate file, thus simplifying the review process and enhancing the debugging experience.

 

Introduction:

 

SQLPackage is a widely used tool for managing SQL Server databases. Its diagnostic logs are crucial for debugging, but their size can make error extraction tedious and time-consuming. This is especially true in support cases where logs can grow to millions of rows. This article presents a PowerShell script that automates the error extraction process, providing a practical solution to this common challenge.

 

The PowerShell Script:

 

The PowerShell script we are introducing reads a SQLPackage diagnostic log file line by line, identifies error messages based on specific patterns, and then saves these error messages to a separate output file. The script is efficient as it doesn’t load the entire file into memory and is user-friendly with clear output summaries and progress updates.

 

param (
    [Parameter(Mandatory=$false)]
    [string]$logFilePath = "C:\path\to\diagnostic.log",
    
    [Parameter(Mandatory=$false)]
    [string]$outFilePath = "C:\path\to\output.txt"
)

# Check if input file exists
if (-Not (Test-Path $logFilePath)) {
    Write-Error "The specified input file does not exist."
    exit
}

# If output file exists, delete it
if (Test-Path $outFilePath) {
    Remove-Item $outFilePath
}

$informationCount = 0
$warningCount = 0
$errorCount = 0
$verboseCount = 0
$lines = 0
$capturingError = $false

try {
    $reader = [System.IO.StreamReader]::new($logFilePath)
    try {
        while ($reader.EndOfStream -eq $false) {
            $line = $reader.ReadLine()
            $lines++
            if ($lines % 2000 -eq 0) {
                Write-Output "Line: $lines"
            }
            
            switch -Regex ($line) {
                "Microsoft.Data.Tools.Diagnostics.Tracer Information:" {
                    $informationCount++
                    $capturingError = $false
                }
                "Microsoft.Data.Tools.Diagnostics.Tracer Warning:" {
                    $warningCount++
                    $capturingError = $false
                }
                "Microsoft.Data.Tools.Diagnostics.Tracer Verbose:" {
                    $verboseCount++
                    $capturingError = $false
                }
                "Microsoft.Data.Tools.Diagnostics.Tracer Error:" {
                    $errorCount++
                    $capturingError = $true
                }
            }

            if ($capturingError) {
                try {
                    Add-Content -Path $outFilePath -Value "Line:$lines Found:$line"
                } catch {
                    Write-Error "Unable to write to the output file."
                    exit
                }
            }
        }
    } finally {
        $reader.Close()
    }
} catch {
    Write-Error "Unable to open the input file."
    exit
}

Write-Output "Summary:"
Write-Output "Input File : $logFilePath"
Write-Output "Output File: $outFilePath"
Write-Output "Lines      : $( "{0:N0}" -f $lines)"
Write-Output "Information: $( "{0:N0}" -f $informationCount)"
Write-Output "Warnings   : $( "{0:N0}" -f $warningCount)"
Write-Output "Errors     : $( "{0:N0}" -f $errorCount)"
Write-Output "Verbose    : $( "{0:N0}" -f $verboseCount)"

 

Conclusion:

 

The provided PowerShell script offers an efficient and practical solution for extracting error messages from massive SQLPackage diagnostic logs. It caters specifically to support cases where logs can be incredibly large, ensuring that database administrators and developers can save time and effort while effectively debugging and troubleshooting issues.

Version history
Last update:
‎Oct 20 2023 10:00 AM
Updated by: