Skip to content

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

Your vault is empty.

License Error MapiExceptionPartialCompletion

MapiExceptionPartialCompletion: public folder migration stalls part way through

10 min read Updated October 5, 2026 Exchange Server

Fix it now

The operation finished for some of the work and not the rest, so the public folder batch will not complete. It is not connectivity and not licensing. Find the request that failed and the folder behind it, rather than restarting the batch and waiting for the same result: Exchange allows only one migration batch for a public folder migration, so there is no clean second attempt to fall back on.

Run the first three in Exchange Online PowerShell and the last on the source server

Get-MigrationBatch | Format-List Identity,Status,TotalCount,SyncedCount,FailedCount
Get-MigrationUser -BatchId <batch> | Where-Object {$_.Status -ne "Synced"}
Get-PublicFolderMailboxMigrationRequest | Get-PublicFolderMailboxMigrationRequestStatistics -IncludeReport | Format-List Status,Message
Get-PublicFolderStatistics | Sort-Object ItemCount -Descending | Select-Object -First 20 Name,ItemCount,TotalItemSize
  1. Read the failing request’s message, not the batch status. The batch only tells you that something did not finish; the request tells you which target mailbox and roughly where.
  2. Correlate that target mailbox against the section of the hierarchy your mapping file assigned to it, then look at those folders on the source.
  3. Check the biggest offenders first. Microsoft recommends removing content from any single public folder larger than 25 GB before migrating it.
  4. Resume the individual request once you have dealt with the folder, and complete the batch only when every migration user reports a synced state.

Check \NON_IPM_SUBTREE\DUMPSTER_ROOT for folder count as well as size. Microsoft documents that more than 10,000 folders under it can cause the migration to fail, and nothing in the batch status hints at it.

If the remaining requests complete and the batch reaches a completed status, you are done. If the same request keeps failing, the next section explains what actually stalls a folder.

Why it happens

A public folder migration is not one operation. Before it starts, two scripts do the planning: Export-ModernPublicFolderStatistics.ps1 reads the source hierarchy, and ModernPublicFolderToMailboxMapGenerator.ps1 turns those statistics into a mapping file that splits the tree across target public folder mailboxes according to a size you choose. Each target mailbox then gets its own migration request. Those requests run in parallel and report back to the batch, and the batch is only as complete as its least healthy request.

So one folder that cannot be copied holds the whole thing open. The two things that stall a folder are size and content. A folder with an enormous item count takes longer to copy than the operation’s time allowance, so the attempt is abandoned and reported as a timeout, repeatedly. A folder containing something the target will not accept produces an invalid parameter result instead: a property that cannot be represented, a permission entry pointing at an account that no longer resolves, or a name the target treats as illegal. The first fails slowly, the second fails almost immediately, and that timing is the cheapest way to tell them apart.

The constraint that shapes everything else is that Exchange allows creating only one migration batch for a public folder migration. There is no partial re-run and no second batch for the stragglers. Whatever the mapping assigned is what the migration will attempt, which is why the sizing decisions made before the batch started matter more here than in a mailbox migration, and why the answer to a stalled folder is nearly always to fix the folder rather than to re-plan around it.

One folder is far larger than the rest

You have this one if The same folder appears in every failed attempt, and its item count or size on the source dwarfs everything around it.

  1. Confirm it with Get-PublicFolderStatistics | Sort-Object ItemCount -Descending. Microsoft recommends deleting content from any single folder over 25 GB before migrating.
  2. Check whether the count is inflated by deleted items still held in the folder’s recoverable items; purging those legitimately reduces the work.
  3. Split the folder into subfolders on the source. It is slower to arrange than a retry and far more reliable.
  4. Resume the request for the target mailbox that folder was mapped to.

The dumpster holds more folders than the migration can handle

You have this one if A request fails without naming an obvious large folder, and the source has a long history of deletions.

  1. Count the folders under \NON_IPM_SUBTREE\DUMPSTER_ROOT on the source.
  2. Microsoft documents that more than 10,000 folders there can cause the migration to fail.
  3. Reduce the count before retrying rather than looking for a folder in the visible tree that is not the problem.

This one is invisible from the batch status and from the folder tree users see, which is why it costs people days.

A folder contains something the target rejects

You have this one if An invalid parameter result rather than a timeout, always the same folder, failing almost immediately.

  1. Inspect the folder name for characters legal on the source but not accepted by the target.
  2. Read its permissions with Get-PublicFolderClientPermission and remove entries referencing accounts that no longer resolve.
  3. Check whether the folder is mail enabled and whether its mail object is intact.
  4. Rename or repair on the source, then resume the request for that mailbox.

The mapping put too much into one target mailbox

You have this one if One target public folder mailbox fails while its siblings complete, and the mapping gave it a disproportionate share.

  1. Compare what the mapping assigned to each target mailbox against the source statistics.
  2. Microsoft’s ceiling is 100 GB per target public folder mailbox, and the recommendation is to plan at roughly half that to leave room for growth.
  3. If the mapping is genuinely unworkable, the migration has to be planned again from the statistics rather than patched with a second batch.

Keep the mapping file with your migration records. If anything has to be redone, it is the document that tells you what was supposed to go where.

Damaged items inside an otherwise healthy folder

You have this one if Skipped item counts rise steadily and the request eventually fails on tolerance rather than on time.

  1. List what is being skipped with Get-MigrationUserStatistics -Identity <user> -IncludeSkippedItems and save the output.
  2. On an on-premises target you can set a tolerance with Set-MigrationBatch -Identity <batch> -BadItemLimit <n>; that parameter is on-premises only and is deprecated from the cloud service, so it is not available for a migration landing in Exchange Online.
  3. Tell the folder owners what was left behind before the source is decommissioned.

Shared folders often hold the only copy of something. A tolerance on a public folder batch abandons content nobody has a personal copy of.

Full reference

The numbers Microsoft publishes for a public folder migration

Limit Value
Maximum size of a target public folder mailbox 100 GB, with roughly half that recommended for planning
Recommended recoverable items quota per target mailbox 15 GB or less
Single source folder worth removing content from Larger than 25 GB
Maximum target public folder mailboxes in the native migration 100, though up to 1000 can be created after migration
Folders under \NON_IPM_SUBTREE\DUMPSTER_ROOT More than 10,000 can cause the migration to fail
Largest total migration Microsoft recommends 5 TB
Migration batches allowed One

Reading the failure type before reaching for a limit

What the report shows What to do about it
A timeout against one large folder, repeatedly Reduce the folder: purge recoverable items, delete content, or split it into subfolders
An invalid parameter result on a specific folder Inspect that folder’s name, properties and permissions on the source
Skipped items counted but the request still running Ordinary damaged items; decide a tolerance you can justify, if the target allows one
One target mailbox failing while others complete The mapping overloaded it; check it against the 100 GB ceiling
Every request failing at the same point Not a folder problem. Look at the endpoint and credentials

Raising a bad item tolerance is the wrong instrument for a folder that is too big to copy inside one attempt. Nothing is being skipped there; the work simply is not finishing.

Cmdlets worth having open

Command What it gives you
Get-MigrationBatch Batch-level status and the synced and failed counts
Get-MigrationUser -BatchId <batch> One row per migration user, which is where the failure shows
Get-PublicFolderMailboxMigrationRequest The underlying requests, one per target public folder mailbox
Get-PublicFolderMailboxMigrationRequestStatistics -IncludeReport The report for a request, which names what failed
Get-PublicFolderStatistics Item counts and sizes on the source, for finding the offender
Complete-MigrationBatch <name> Finishes the migration once every user reports synced

What completion actually closes

Once a public folder migration batch is completed, no further data can be synchronised from the on-premises servers. That makes completion a one-way door rather than a milestone, and it is the reason to be pedantic about every request reporting a synced state before you press it. A folder that never copied is a folder that will need moving by hand afterwards, from a source you may already have scheduled for decommissioning.

When a licence is the actual fix

Nobody should buy anything to unstick a batch. Every step above is planning and configuration, and the failure is in the data rather than in your entitlement. The licences matter at the destination: a public folder migration is usually the last step in retiring an on-premises Exchange server, and every mailbox that lands in Microsoft 365 needs a plan that includes Exchange Online. Microsoft 365 Business Standard is the common choice for a smaller organisation finishing this work. Settle it before the batch runs, because completion closes synchronisation from the on-premises servers permanently and is a poor moment to discover a purchasing conversation. We can check which plan covers the storage and archive requirements the migrated content will need.

Every code this article covers

Code What it points at Source
MapiExceptionPartialCompletion The operation succeeded for part of the work and failed for the rest, leaving the batch short of complete not published by the vendor
MapiExceptionTimeout The operation exceeded its time allowance, normally a folder too large to copy in one attempt not published by the vendor
MapiExceptionInvalidParameter Something in the folder could not be represented in the target, such as a name, property or permission entry not published by the vendor

Confirm the fix worked

  1. Get-MigrationUser -BatchId <batch> reports every entry as synced.
  2. Folder counts in the target match the source for the folders that previously failed.
  3. The batch reaches a completed status rather than a partially completed one.
  4. The migrated folders open from a client with both content and permissions intact.
  5. The mapping file and the source statistics are saved with the migration records.

Questions people ask about this

Should I restart the whole batch?

You cannot, in the sense people usually mean. Exchange allows only one migration batch for a public folder migration, so there is no second attempt to fall back on. Resume the specific request that failed once you have dealt with the folder.

How do I find the folder from the error alone?

You usually cannot. Read the request report rather than the batch status, and correlate the failing target mailbox with the section of the hierarchy your mapping assigned to it.

Is there a supported way to migrate a folder that is simply too big?

Reduce it. Purge the recoverable items, delete content, or split it into subfolders on the source. Microsoft recommends removing content from any single folder larger than 25 GB before migrating, which is the same advice from the other direction.

Can I raise the bad item limit on a public folder batch?

On an on-premises target, yes. The BadItemLimit parameter on Set-MigrationBatch is available only in on-premises Exchange and is deprecated from the cloud service, so for a migration into Exchange Online it is not the lever you have.

Does any of this require buying something?

No. Every step here is configuration and planning. Licences matter for the mailboxes that live in Microsoft 365 afterwards, not for finishing a stalled batch.

Related error codes

Was this article helpful?

Your feedback helps us improve our documentation.

Related articles

Free Fix 0x80090327 and 0x800B0109: Exchange TLS fails on certificate validation Free Fix 0x80070643 during an Exchange CU: setup fails and rolls the server back License Error Event ID 2601: Exchange cannot contact a domain controller or global catalog License Error Event ID 1069 and Event ID 1146: cluster faults that take a whole DAG offline
โ† Back to Knowledge Base