DAG Replication Timeout & Corruption - Fix Guide 2025

# Check replication health
Get-MailboxDatabaseCopyStatus -Identity "DB01" | Format-List Name,Status,
  CopyQueueLength,ReplayQueueLength,ContentIndexState,LastInspectedLogTime

# Key metrics to monitor:
# CopyQueueLength: Logs waiting to be copied (should be 0-2)-2)
# ReplayQueueLength: Logs waiting to be replayed (should be < 10)
# ContentIndexState: Should be "Healthy"
# Status: Should be "Mounted" (active) or "Healthy" (passive)"Healthy" (passive)

# Check current status
$dbCopy = "DB01\EXCH02"
Get-MailboxDatabaseCopyStatus -Identity $dbCopy | Format-List Status,*Queue*,LastInspectedLogTime

# Resume the copy
Resume-MailboxDatabaseCopy -Identity $dbCopy

# Monitor replication progress (run every 30 seconds)
while ($true) {
    $status = Get-MailboxDatabaseCopyStatus -Identity $dbCopy
    Write-Host "$(Get-Date -Format 'HH:mm:ss') - Copy: $($status.CopyQueueLength) | Replay: $($status.ReplayQueueLength) | Status: $($status.Status)"-Format 'HH:mm:ss') - Copy: $($status.CopyQueueLength) | Replay: $($status.ReplayQueueLength) | Status: $($status.Status)"
    if ($status.CopyQueueLength -eq 0 -and $status.Status -eq "Healthy") {
        Write-Host "Replication caught up!" -ForegroundColor Green
        break
    }
    Start-Sleep -Seconds 30
}

# Step 1: Suspend the problematic copy
$dbCopy = "DB01\EXCH02"
Suspend-MailboxDatabaseCopy -Identity $dbCopy -Confirm:$false

# Step 2: Remove existing database files (optional, speeds up reseed)
# WARNING: This deletes the local copy - ensure other copies exist!
# Run this on the target server (EXCH02)
$dbPath = (Get-MailboxDatabase DB01).EdbFilePath.PathName
$logPath = (Get-MailboxDatabase DB01).LogFolderPath.PathName
# Remove-Item "$dbPath" -Force"$dbPath" -Force
# Remove-Item "$logPath\*.log" -Force"$logPath\*.log" -Force

# Step 3: Start reseed
Update-MailboxDatabaseCopy -Identity $dbCopy -DeleteExistingFiles

# Step 4: Monitor reseed progress
while ($true) {
    $status = Get-MailboxDatabaseCopyStatus -Identity $dbCopy
    Write-Host "$(Get-Date -Format 'HH:mm:ss') - Status: $($status.Status) | $($status.SeedingProgress)%"-Format 'HH:mm:ss') - Status: $($status.Status) | $($status.SeedingProgress)%"
    if ($status.Status -eq "Healthy") {
        Write-Host "Reseed complete!" -ForegroundColor Green
        break
    }
    Start-Sleep -Seconds 60
}

# Use a specific server as seed source (useful when active copy is busy)
$dbCopy = "DB01\EXCH02"
$sourceServer = "EXCH03"  # Another healthy passive copy

Update-MailboxDatabaseCopy -Identity $dbCopy -SourceServer $sourceServer -DeleteExistingFiles

# Or use a specific network for faster seeding
Update-MailboxDatabaseCopy -Identity $dbCopy -Network "DAGNetwork02" -DeleteExistingFiles

# Reseed only the search catalog (much faster than full reseed)
$dbCopy = "DB01\EXCH02"

Update-MailboxDatabaseCopy -Identity $dbCopy -CatalogOnly

# Monitor content index status
Get-MailboxDatabaseCopyStatus -Identity $dbCopy | Select-Object ContentIndexState,ContentIndexErrorMessage

# Check current DAG network configuration
Get-DatabaseAvailabilityGroupNetwork -Identity DAG01 | Format-List Name,Subnets,ReplicationEnabled

# Disable client traffic on replication network
Set-DatabaseAvailabilityGroupNetwork -Identity "DAG01\ReplicationNetwork" -ReplicationEnabled $true -IgnoreNetwork $false

# Enable compression for WAN replication
Set-DatabaseAvailabilityGroup -Identity DAG01 -NetworkCompression Enabled

# Enable encryption for secure replication
Set-DatabaseAvailabilityGroup -Identity DAG01 -NetworkEncryption Enabled

# 1. Check all database copy status
Get-MailboxDatabaseCopyStatus * | Format-Table DatabaseName,MailboxServer,Status,CopyQueueLength,ReplayQueueLength,ContentIndexState

# 2. Run full replication health test
Test-ReplicationHealth | Format-Table Server, Check, Result -AutoSize

# 3. Verify no pending failures
Get-MailboxDatabaseCopyStatus * | Where-Object {$_.Status -notmatch "Mounted|Healthy"}

# 4. Test failover capability (optional - causes brief disruption)
# Move-ActiveMailboxDatabase -Identity DB01 -ActivateOnServer EXCH02 -Confirm:$false-Identity DB01 -ActivateOnServer EXCH02 -Confirm:$false
# Then move back:
# Move-ActiveMailboxDatabase -Identity DB01 -ActivateOnServer EXCH01 -Confirm:$false-Identity DB01 -ActivateOnServer EXCH01 -Confirm:$false

# 5. Check event logs for any new errors
Get-EventLog -LogName Application -Source "MSExchangeRepl" -Newest 20 |
    Where-Object {$_.EntryType -eq "Error"}

# Set up scheduled monitoring (run every 5 minutes)
$threshold = 10  # Alert if copy queue exceeds this

$badCopies = Get-MailboxDatabaseCopyStatus * | Where-Object {
    $_.CopyQueueLength -gt $threshold -or
    $_.Status -notmatch "Mounted|Healthy" -or
    $_.ContentIndexState -ne "Healthy"
}

if ($badCopies) {
    $body = $badCopies | Format-Table DatabaseName,MailboxServer,Status,CopyQueueLength | Out-String
    Send-MailMessage -To "admin@company.com" -Subject "DAG Replication Alert" `
        -Body $body -SmtpServer "mail.company.com"
}

# Comprehensive DAG health check
Write-Host "=== DAG Health Report ===" -ForegroundColor Cyan

# Check DAG configuration
Get-DatabaseAvailabilityGroup | Format-List Name,WitnessServer,WitnessDirectory

# Check all members
Get-DatabaseAvailabilityGroupServer | Format-Table Name,DatabaseCopyAutoActivationPolicy

# Run replication health
Test-ReplicationHealth | Format-Table Server,Check,Result

# Check for long replay queues
Get-MailboxDatabaseCopyStatus * | Where-Object {$_.ReplayQueueLength -gt 100}

# Verify cluster health
Get-ClusterNode | Format-Table Name,State