Skip to content

Est. 2011ยทMicrosoft Partner ID 7033487ยทDelivery under 3 minยทSupport 7 days a week

Your vault is empty.

License Error 452 4.3.1

452 4.3.1 insufficient system resources: Exchange back pressure explained

10 min read Updated October 5, 2026 Exchange Server

Fix it now

Microsoft publishes 452 4.3.1 as insufficient system resources, and names two causes: free disk space is low, or available memory is low. Transport is refusing work it cannot safely queue. Find the resource that is under pressure and relieve it; restarting the service buys minutes.

Run these in the Exchange Management Shell on the affected server

Get-Queue | Format-Table Identity,DeliveryType,Status,MessageCount
Get-Queue | Where-Object {$_.MessageCount -gt 500}
Get-EventLog -LogName Application -Source MSExchangeTransport -Newest 50 | Where-Object {$_.EventID -in 15004,15005,15006,15007}
  1. Read the 15004 entry. It names the resource and reports that pressure increased from one utilisation level to another; the matching decrease is 15005, and that is the event you are waiting to see afterwards.
  2. Check free space on the volume holding the queue database and on the volume holding its transaction logs. This is the cause more often than anything else.
  3. Check memory: the EdgeTransport.exe process on its own, and the server overall. Both are monitored separately.
  4. Free the resource that is actually under pressure. Remove log files that are genuinely logs, never the queue database or its transaction logs.

Do not raise the thresholds to make the message stop. They exist to stop transport damaging its own queue database, and the failure you get instead is worse than a temporary rejection.

If mail is being accepted and 15005 has recorded the resource returning to a lower level, you are done. If not, the next section explains what transport measures and what it does at each level.

Why it happens

Back pressure is a protection mechanism, not a fault. The transport service continuously measures a set of resources and reports each as normal, medium or high utilisation. As pressure rises it sheds work in stages rather than failing outright, which is why the sender sees a temporary rejection and the message is not lost.

Microsoft documents seven monitored resources, and the list is worth knowing because people routinely check only two of them: DatabaseUsedSpace, the disk holding the queue database; UsedDiskSpace for the queue database transaction logs; UsedDiskSpace for content conversion; PrivateBytes, the memory used by EdgeTransport.exe; SystemMemory, the memory used by all processes; UsedVersionBuckets, uncommitted queue database transactions held in memory; and QueueLength for the submission queue.

At medium utilisation Microsoft’s documented behaviour is to reject incoming messages from non-Exchange servers, reject submissions from the Pickup and Replay directories, pause message resubmission, and have shadow redundancy reject messages. At high utilisation it adds everything above plus rejecting incoming messages from other Exchange servers and refusing submissions from mailbox databases. That staging is why the symptom often looks partial: external mail bounces while internal mail keeps moving, for a while.

The volume holding the queue database or its logs is nearly full

You have this one if Free space on that volume is low and Event ID 15006 has been logged, which is transport saying it is rejecting messages because available disk space is below the configured threshold.

  1. Find what is consuming the space before deleting anything. Exchange writes protocol, connectivity and diagnostic logging continuously under %ExchangeInstallPath%Logging and %ExchangeInstallPath%TransportRoles\Logs.
  2. Remove old files that are genuinely logs. Never remove the queue database or its transaction logs by hand.
  3. Check the IIS logs on the same volume. They are frequently the largest single consumer and nothing cleans them up for you.
  4. Watch for Event ID 15005 recording the pressure decreasing before you declare it fixed.

If the volume fills repeatedly, set up log housekeeping and alerting rather than clearing it by hand each time. This failure is entirely predictable and therefore entirely preventable.

Transport is short of memory

You have this one if Event ID 15007 appears, which is transport rejecting message submissions because it continues to consume more memory than the configured threshold.

  1. Note that PrivateBytes and SystemMemory are separate resources. One is the transport process, the other is the whole server, and they can trip independently.
  2. Check what else is running. Exchange expects to have the machine largely to itself.
  3. Microsoft’s guidance for the 4.3.1 memory case names dynamic memory on a virtual machine specifically: use static memory on Exchange virtual machines.
  4. Restart the transport service only to regain control, and treat that as buying time rather than as a repair.

A mail loop or bulk send has flooded the submission queue

You have this one if One queue holds a very large number of messages, the volume appeared suddenly, and the messages share a sender, a subject or a recipient.

  1. Identify the source: Get-Message -Queue '<server>\<queue>' -ResultSize 100 | Format-Table FromAddress,Recipients,Subject,Status.
  2. Suspend the offending queue so the rest of mail flow can drain: Suspend-Queue -Identity '<server>\<queue>'.
  3. Fix the cause. A forwarding loop, a compromised account or an application stuck in a retry cycle are the usual candidates.
  4. Remove the offending messages deliberately once you know what they are, then resume the queue.

It is not back pressure at all: 421 4.3.2

You have this one if The sending server reports 421 4.3.2 service not available or service not active rather than 452 4.3.1.

  1. Microsoft publishes three causes for 4.3.2, and none of them is back pressure. Check them in order.
  2. Look for a custom receive connector listening on port 25 in the Transport service on a Mailbox server. Custom connectors on port 25 belong in the Front End Transport service.
  3. Check for inactive server components: Get-ServerComponentState -Identity <ServerName>, and reactivate with Set-ServerComponentState -Identity <ServerName> -Component ServerWideOffline -State Active -Requester Maintenance.
  4. Check for incompatible transport agents, particularly after an Exchange update. Identify the agent, then disable or uninstall it.

This is the correction that saves the most time on this error. Reading 4.3.2 as back pressure sends you looking for disk space on a server whose components are simply still in maintenance mode after a patch window.

Full reference

The events transport writes, and which one you are waiting for

Event ID Type What it records
15004 Error Resource pressure increased from the previous utilisation level to the current one
15005 Information Resource pressure decreased from the previous utilisation level to the current one
15006 Error Transport is rejecting messages because available disk space is below the configured threshold
15007 Error Transport is rejecting message submissions because it continues to consume more memory than the configured threshold

All four have source MSExchangeTransport and category Resource Manager. The pairing matters: 15004 is the alarm and 15005 is the all-clear, so an incident that shows a 15004 with no later 15005 has not been resolved, whatever the queues currently look like.

Where the thresholds live

Back pressure settings are held in the transport service’s application configuration file, %ExchangeInstallPath%Bin\EdgeTransport.exe.config. It is XML, and every monitored resource has its own entries there. Knowing where the file is matters mainly so you can confirm nobody has already edited it: a server that stops reporting back pressure after someone raised a threshold is not a healthy server, it is a quiet one.

The transport queue database holds messages that have been accepted but not yet delivered. Deleting it discards that mail permanently, with no notification to the senders. If you must move or rebuild it, stop the transport service cleanly, take a copy first, and be certain what you are discarding.

Matching the observation to the resource

Observation Resource
Low free space where the queue database lives DatabaseUsedSpace
Low free space where its transaction logs live UsedDiskSpace for the queue database logging path
Low free space under the transport data folder UsedDiskSpace for content conversion
EdgeTransport.exe memory climbing steadily PrivateBytes
Server-wide memory exhausted, transport not unusual SystemMemory
A single queue with an enormous message count QueueLength for the submission queue
Pressure appears under load and clears when quiet The server is undersized for its peak, not broken

When it never clears

  • Pressure that persists while the server is idle means something is holding the resource permanently. Logs nobody removes are the usual answer.
  • Confirm you are looking at the right volume. The queue database and its logs can sit on different disks from the mailbox databases and from the operating system.
  • If verbose protocol or connectivity logging was turned on for troubleshooting and never turned off, that alone will fill a volume on a busy server within weeks.
  • Check whether the queue database has been relocated. Microsoft documents moving it to a different disk as the response to a persistently tight volume.

When a licence is the actual fix

Freeing disk space and clearing a mail loop cost nothing, and if that is what you are facing then stop there. The licensing question arises only when the honest answer is that the server is too small for the job: not enough memory, no room to give the transport queue its own volume, or a machine sharing roles it should never have carried. Microsoft’s own guidance for the memory case is to use static rather than dynamic memory on Exchange virtual machines, which sometimes means the host has to grow rather than the guest. Rebuilding onto a properly sized server means a Windows Server licence and the client access licences to go with it, and Arco supplies Windows Server 2025 Standard along with the CALs. Get the sizing right first, then buy to that, rather than the other way round.

Every code this article covers

Code What it points at Source
452 4.3.1 Insufficient system resources. Microsoft’s documented causes are low free disk space, for example on the disk holding the queue database, and low available memory Microsoft Learn
421 4.3.2 Service not available, or service not active. Microsoft’s documented causes are a custom receive connector on port 25 in the Transport service, inactive Exchange server components, and incompatible transport agents. It is not a back pressure code Microsoft Learn
Event ID 15004 Resource pressure increased from the previous utilisation level to the current one. Source MSExchangeTransport, category Resource Manager Microsoft Learn
Event ID 15006 The Microsoft Exchange Transport service is rejecting messages because available disk space is below the configured threshold Microsoft Learn

Confirm the fix worked

  1. Event ID 15005 records the resource pressure decreasing after your change.
  2. Free space on the queue database and log volumes is well clear of the threshold, with headroom for a busy day.
  3. Get-Queue shows queues draining and no queue with an unexplained message count.
  4. An external sender delivers a test message immediately, without a temporary rejection.

Questions people ask about this

Will mail be lost while this is happening?

No. The failure is temporary, so sending servers hold the mail and retry. Mail is lost only if the condition outlives the sender’s own retry lifetime, which gives you hours rather than minutes, but not days.

Should I raise the thresholds to make it stop?

Almost never. The thresholds exist to stop transport damaging its own queue database. Raising them removes the warning without adding any capacity.

I am getting 421 4.3.2, not 452 4.3.1. Is that the same thing?

No, and this is where most of the wasted time on this error goes. Microsoft publishes 4.3.2 as service not available or service not active, with three causes: a custom receive connector on port 25 in the Transport service, inactive server components, and incompatible transport agents. Check Get-ServerComponentState before you look at disk space.

Do I need a bigger licence to fix back pressure?

No licence changes how much disk or memory a server has. If the fix is to free space or stop a loop, that is free. Licensing enters only if you conclude the server needs replacing.

Why did it start after we enabled extra logging?

Because verbose protocol and connectivity logging generates a great deal of data on a busy server. Logging turned on for troubleshooting and never turned off is a classic cause of a volume filling weeks later.

Related error codes

Was this article helpful?

Your feedback helps us improve our documentation.

Related articles

Free Fix HTTP 500 on Microsoft-Server-ActiveSync: every mobile device stops syncing License Error Event ID 226 and Event ID 9519: an Exchange database will not start License Error Error 1642: an Exchange security update refuses to patch this build License Error 552 5.2.2 mailbox full: recipient quota and hourly receive limits in Exchange
โ† Back to Knowledge Base