Exchange Server VSS backup error codes: writer states, VSS_E codes and first fixes

When an Exchange backup fails, the product's message usually hides a Volume Shadow Copy Service (VSS) code such as 0x800423F3 or a writer stuck in a failed state. This quick reference covers 30 of them: the VSS HRESULTs, the writer states and last errors that vssadmin list writers shows, and the Exchange and Windows events that go with them, each with the usual cause on Exchange Server SE, 2019 and 2016 and the first check. Press Ctrl+F and type your code.

How to read these codes#

Exchange Server supports only Exchange-aware, VSS-based backups. The backup product (the requester) asks VSS for a snapshot, the Microsoft Exchange Writer, hosted by the Microsoft Exchange Replication service (MSExchangeRepl), prepares the databases, and a provider takes the snapshot. A failure in any of the three becomes an HRESULT:

• 0x800423F1 to 0x800423F4: writer errors: a writer (on an Exchange server, usually the Microsoft Exchange Writer) failed.
• Other 0x800423xx codes: VSS and provider errors.
• Negative codes such as -2402: the database engine, covered in the ESE error codes quick reference.

You'll find them in the backup job log, in vssadmin list writers (State and Last error), in Microsoft's BETEST tool (Status for writer Microsoft Exchange Writer: STABLE (0x800423f3 - VSS_E_WRITERERROR_RETRYABLE)) and in the Application log under the VSS, MSExchangeRepl and ESE sources. Some products print the Windows message text instead, such as The writer experienced a non-transient error. If the backup process is retried, the error is likely to reoccur. (0x800423F4); some log the code in decimal. Product-specific messages are in the guides for Veeam (and Veeam Explorer restores), Backup Exec and NetBackup and Commvault.

# Elevated prompt on the server that runs the backup (in a DAG, the passive copy's server)
vssadmin list writers | Select-String 'Microsoft Exchange Writer' -Context 0,4
vssadmin list providers
vssadmin list shadowstorage

'{0:X8}' -f 2155348129     # 807800A1  (decimal code from a log)
'{0:X8}' -f -2147212301    # 800423F3  (negative decimal)

A healthy writer looks like this before and after every job:

Writer name: 'Microsoft Exchange Writer'
   Writer Id: {...}
   Writer Instance Id: {...}
   State: [1] Stable
   Last error: No error

Writer errors#

Check in vssadmin which writer failed before blaming Exchange. A writer also vetoes a snapshot set that lacks a volume it needs, so always back up each database together with its log volume. Microsoft's GetWriterStatus reference describes these four errors by name only. The hex values are the ones tools print: Microsoft's own pages show 0x800423f3 in BETEST output and 0x800423F4 in a Windows Server Backup failure, and 0x800423F1 and 0x800423F2 are shown as backup products report them.

For retryable errors on passive-copy backups, Microsoft's guidance is to keep KeepAliveTime on the Exchange servers below the idle-session timeout of network devices (and above 15 minutes), back up passive copies rather than active ones, never back up both at once, avoid moving databases during the job and, for busy servers, raise the log-range query timeout from its 30-second default for better tolerance of network or performance problems:

New-ItemProperty -Path 'HKLM:\SOFTWARE\Microsoft\ExchangeServer\v15\Replay\Parameters' -Name QueryLogRangeTimeoutInMsec -PropertyType DWord -Value 120000
Restart-Service MSExchangeRepl    # in a quiet period

Provider and snapshot codes#

These come from VSS or the provider. Hex values are as listed in Microsoft's AddToSnapshotSet reference, except 0x80042318 and 0x80042319, which Microsoft's pages name without a value (shown here as backup products report them). The remaining codes are listed by name only.

Writer states and last errors in vssadmin#

vssadmin list writers prints the state number in brackets. The numbers come from Microsoft's VSS_WRITER_STATE enumeration: 6 to 15 are failure states and all display as Failed, so the number tells you which phase failed.

Exchange and Windows backup events#

How to verify a fix#

1. Writer: State: [1] Stable and Last error: No error before the job, and again after it, not stuck at [5].
2. Job: the backup completes without VSS warnings in its log.
3. Exchange recorded it: LastFullBackup or LastIncrementalBackup shows the new time. Copy and differential backups never truncate logs, by design.
4. Logs truncate: after a full or incremental backup the log count drops once every copy has replayed them; lagged copies keep theirs.
5. Quiet event log: no new VSS, MSExchangeRepl or ESE errors in the job window.
6. Control test: if the product still fails while a Windows Server Backup VSS full backup of the same volumes succeeds, look at the product's provider and settings.

Get-MailboxDatabase -Status | Format-Table Name,LastFullBackup,LastIncrementalBackup -AutoSize
Get-MailboxDatabaseCopyStatus * | Format-Table Name,Status,CopyQueueLength,ReplayQueueLength -AutoSize
Get-WinEvent -FilterHashtable @{LogName='Application'; ProviderName=@('VSS','MSExchangeRepl','ESE'); Level=2; StartTime=(Get-Date).AddHours(-12)} |
    Format-Table TimeCreated,ProviderName,Id -AutoSize

References#

• VSS_E_WRITERERROR_RETRYABLE error when backing up a passive database copy in an Exchange DAG (Microsoft Learn)
• IVssBackupComponents::DoSnapshotSet: VSS return codes and meanings (Microsoft Learn)
• IVssBackupComponents::AddToSnapshotSet: return codes with hex values (Microsoft Learn)
• IVssBackupComponents::GetWriterStatus: writer failure codes (Microsoft Learn)