Understanding & Troubleshooting Exchange Server Back Pressure (Event ID 15004 / 15006 & Get-ExchangeDiagnosticInfo)

When an Exchange Server 2016, Exchange Server 2019, or Exchange Server Subscription Edition (SE) Mailbox or Edge Transport server begins rejecting incoming messages with 452 4.3.1 Insufficient system resources and logs MSExchangeTransport Event ID 15004 or 15006 in the Application Event Log, the server has entered Back Pressure.

Back Pressure is a self-protection mechanism built into the Microsoft Exchange Transport service (EdgeTransport.exe). It continuously monitors disk free space, memory utilization, and uncommitted ESE version buckets in mail.que, throttling or halting mail acceptance before the server suffers catastrophic disk exhaustion or database corruption.

In this guide, we explain how the three Back Pressure levels (Normal, Medium, and High) work, how to query live resource utilization in XML using Get-ExchangeDiagnosticInfo, and how to permanently resolve disk-based Back Pressure by moving mail.que to a dedicated volume via Move-TransportDatabase.ps1.


Architecture: Exchange Transport Back Pressure Thresholds & Responses

Exchange Server Transport Resource Monitoring & Back Pressure States
State 1: Low / Normal
Full Mail Flow
  • All monitored disk & memory metrics below threshold
  • Accepts all internal & external MAIL FROM
  • Logs Event ID 15005 upon recovery
State 2: Medium
Selective Throttling
  • Logs Event ID 15004 (Resource pressure increased)
  • Authoritative (internal) mail still accepted
  • External anonymous mail tarpitted or rejected
State 3: High
Complete Mail Halt
  • Logs Event ID 15006 (Severe disk/RAM exhaustion)
  • All new MAIL FROM rejected with 452 4.3.1
  • Mailbox Submission paused until pressure drops

System Resources Monitored by Exchange Back Pressure

The Microsoft Exchange Transport service (%ExchangeInstallPath%Bin\EdgeTransport.exe.config) monitors the following seven system resources every few seconds:

Monitored Resource NameWhat It MeasuresCommon Root Cause When Triggered
UsedDiskSpace[%ExchangeInstallPath%TransportRoles\data\Queue]Hard drive utilization on the volume hosting mail.que and transaction logs (trn*.log)IIS logs, ETL traces, or transport logs filling up the C:\ drive
DatabaseUsedSpace[...\TransportRoles\data\Queue]Disk space consumed specifically by the mail.que database minus free disk headroomMassive outbound mail storm or stuck unreachable remote domain queues
UsedDiskSpace[%ExchangeInstallPath%TransportRoles\data]Disk space on the volume used for temporary content conversionLarge attachments being converted simultaneously on a low-space volume
UsedVersionBuckets[...\Queue\mail.que]Uncommitted in-memory ESE database transactions for mail.que (High threshold default: 2500)High Disk I/O latency on the mail.que storage LUN preventing ESE commits
QueueLength[SubmissionQueue]Number of messages waiting in the categorizer Submission queueSlow LDAP/AD queries, broken transport agents, or antivirus scanning bottlenecks
PrivateBytesRAM committed specifically by EdgeTransport.exeMemory leak in a third-party transport agent
SystemMemoryOverall Windows OS physical memory utilization (94% High threshold)Missing msExchMinAdminMemory / pagefile sizing or unconstrained co-located agents

Step 1: Identify Which Exact Resource Triggered Back Pressure (Get-ExchangeDiagnosticInfo)

Instead of guessing whether Back Pressure was triggered by disk space, memory, or mail.que version buckets, run this PowerShell script in the Exchange Management Shell (EMS) to query the live ResourceThrottling component inside EdgeTransport.exe:

POWERSHELL
# Query the live ResourceThrottling XML diagnostics from EdgeTransport
[xml]$Diag = Get-ExchangeDiagnosticInfo -Server $env:COMPUTERNAME -Process EdgeTransport -Component ResourceThrottling

$ResourceTracker = $Diag.Diagnostics.Components.ResourceThrottling.ResourceTracker
Write-Host "=== Overall Transport Back Pressure Level: $($ResourceTracker.maxPressureLevel) ===" -ForegroundColor Cyan

# Parse and display each monitored resource, its current level, and High/Medium/Low thresholds
$ResourceTracker.ResourceMeter | ForEach-Object {
    [PSCustomObject]@{
        Resource              = $_.name
        CurrentPressureLevel  = $_.currentPressureTransitions
        Pressure              = $_.pressure
        LowToMediumThreshold  = $_.PressureTransitions.MediumToLow
        MediumToHighThreshold = $_.PressureTransitions.LowToMedium
        HighThreshold         = $_.PressureTransitions.MediumToHigh
    }
} | Format-Table -AutoSize

Also Check Windows Application Event Log for Event IDs 15004, 15005, 15006, and 15007:

POWERSHELL
Get-WinEvent -FilterHashtable @{
    LogName      = "Application"
    ProviderName = "MSExchangeTransport"
    Id           = @(15004, 15005, 15006, 15007)
} -MaxEvents 10 | Select-Object TimeCreated, Id, Message | Format-List

Step 2: Immediate & Permanent Remediation for Disk-Based Back Pressure

In 90% of enterprise incidents, Back Pressure (Event ID 15006) is triggered because mail.que is located on the default C:\ system partition and IIS logs (C:\inetpub\logs\LogFiles) or Exchange diagnostic traces (%ExchangeInstallPath%Logging) have consumed the remaining free space.

Fix A: Immediately Reclaim Disk Space on C:\ (Emergency Relief)

Purge IIS and Exchange diagnostic .log / .blg files older than 7 days to immediately drop disk utilization below the MediumToLow threshold (Transport will automatically resume mail flow within 60 seconds without a reboot):
POWERSHELL
# Purge IIS logs and Exchange diagnostic logs older than 7 days
$Cutoff = (Get-Date).AddDays(-7)
Get-ChildItem "C:\inetpub\logs\LogFiles", "$env:ExchangeInstallPath\Logging" -Recurse -File -Include *.log, *.blg -ErrorAction SilentlyContinue |
    Where-Object { $_.LastWriteTime -lt $Cutoff } |
    Remove-Item -Force -ErrorAction SilentlyContinue

Fix B: Move mail.que to a Dedicated Disk Drive Using Move-TransportDatabase.ps1 (Permanent Fix)

Microsoft includes a built-in PowerShell script (Move-TransportDatabase.ps1) that automatically calculates optimal free space, updates EdgeTransport.exe.config, and moves the queue database (mail.que) and its transaction logs to a dedicated non-OS volume:
POWERSHELL
# Run from an elevated Exchange Management Shell (EMS)
Set-Location "$env:ExchangeInstallPath\Scripts"

.\Move-TransportDatabase.ps1 `
    -queueDatabasePath "T:\TransportQueue\Queue" `
    -queueDatabaseLoggingPath "T:\TransportQueue\QueueLogs" `
    -iPFilterDatabasePath "T:\TransportQueue\IPFilter" `
    -iPFilterDatabaseLoggingPath "T:\TransportQueue\IPFilterLogs" `
    -temporaryStoragePath "T:\TransportQueue\Temp"