Exchange Jet error -1018 means a database page failed its checksum.
Microsoft ties it to disk, driver, firmware and hardware faults, so check storage first. Replace the bad page from a DAG copy or a backup. Export mail before a hard repair, using a duplicate of the file.
Backups stop with error -1018 and ESE event 474 appears in the Application log. Exchange found a page whose stored checksum does not match its contents.
What you do next decides whether you lose one page of mail or many. Exchange Jet error -1018 is a warning, because Microsoft tells you to plan for imminent failure or more random damage until you find the root cause.
What does Exchange Jet Error -1018 Mean?
Exchange Jet error -1018 is JET_errReadVerifyFailure, which Microsoft's ESE error list describes as a database page with a checksum error, per Extensible Storage Engine Error Codes. In plain terms, Exchange logs a database page checksum error.
Pages in an Exchange database store a checksum. Exchange recalculates it when a page comes back from disk and raises -1018 when the two values differ. Exchange raises -1018 as well when a page number does not match the page position in the file, per ESE 474 -1018: Unrecoverable Error Detected in Database.
ESE logs the failure under three event IDs. Microsoft documents them in its Exchange 2007 and Exchange 2010 management pack pages.
| Event ID | ESE error | What Exchange found |
|---|---|---|
| 474 | -1018 | Page checksum mismatch |
| 475 | -1018 | Page number mismatch |
| 476 | -1019 | Blank page where data should sit |
For Exchange 2010, Microsoft lists four symptoms of event 474, per Database page read failed verification because of a -1018 error (page checksum mismatch). This error is often caused by hardware issues. Users cannot send or receive mail, Outlook does not start on client computers, online backups fail and online defragmentation reports a checksum mismatch.
Microsoft adds that damage to a critical global table can stop a database from starting and leave repair unsuccessful or only partly successful, per ESE 474 -1018: Unrecoverable Error Detected in Database.
What Causes a Database Page Checksum Error?
Storage sits behind -1018. Microsoft's Exchange Server Analyzer article names defective disk subsystem hardware and outdated or incompatible disk drivers or firmware.
Exchange itself can build a page with a wrong checksum. Transient memory errors can send a correct page to the wrong location.
Microsoft's Exchange 2007 guidance puts page level damage behind a page number mismatch on driver, firmware or hardware issues 99 percent of the time, per The database page read failed verification because of a -1018 error (page number mismatch).
| Cause | What Microsoft says | First check |
|---|---|---|
| Disk hardware | Defective disk subsystem parts cause -1018 | Disk vendor diagnostics |
| Drivers and firmware | Outdated or incompatible versions cause -1018 | Update controller BIOS and firmware |
| RAM or firmware fault | Error can be transient while page stays undamaged | System diagnostic tests |
| Antivirus file scans | Scanner locks or quarantines database file and can cause -1018 errors | Exclude .chk, .edb and .log |
| Exchange itself | Builds a wrong checksum or stores a page in the wrong place | System log review |
Exclude .chk, .edb and .log extensions along with your database folders when antivirus scans the server, as Anti-Virus Software in the Operating System on Exchange Servers advises for Exchange 2013.
“A -1018 is a storage problem wearing a database error message. Replace the disk first, or the next repair has fresh damage to deal with.”
How can You Test a Database for Bad Pages?
Dismount the database and use eseutil /k on copies of its files. Microsoft says Exchange 2007 checksum mode verifies the integrity of database pages and needs the database dismounted, per How to Run Eseutil /K (Checksum).
eseutil /k "D:\ExchangeDB\DB01\DB01.edb"
Microsoft says checksum mode does not test single pages in the database. Page dump mode, described in Eseutil /K Checksum Mode, tells you if the checksum of a named page is correct.
Record the errors the check reports before you choose a fix. One or two bad pages call for a different plan than thousands do. Other eseutil switches are covered in Eseutil Commands Reference.
How do You Fix Exchange Jet Error -1018?
Exchange Jet error -1018 fixes start with storage, then a page patch, a mailbox move or a restore. Hard repair comes last.
| Situation | Fix | Microsoft source |
|---|---|---|
| DAG copy is healthy | Page patching replaces corrupted page from another copy | Exchange 2010 page |
| Lagged copy shows -1018 | Reseed with Update-MailboxDatabaseCopy | Exchange 2016 and 2019 page |
| Database mounts | Move mailboxes to a good or new database, then delete corrupted one | Exchange 2000 to 2007 article |
| Known good backup exists | Restore backup | Exchange 2000 to 2007 article |
| No copy and no backup | Export mail first, then hard repair a duplicate | Exchange 2007 repair page |
Microsoft lists three recovery methods in this order for Exchange 2000 to 2007. Move mailboxes first, restore a known good backup second and hard repair last, per ESE 474 -1018: Unrecoverable Error Detected in Database.
Skip both steps if you export mail from the repaired copy and discard that copy afterward. Before You Run Eseutil /p covers the full hard repair process.
Error -1018 sits among the eight mount error codes in Exchange Database Won't Mount.
How do You Save Mail From a Database With Bad Pages?
Before a repair, save mail with Exchange EDB Converter from Univik, which reaches a dirty or dismounted .edb offline under Windows 11 and 10.
Univik Software applies saved E00 logs to a temporary duplicate of your -1018 database. When logs are missing, Univik Software repairs that temporary copy in the app.
Raw scanning reaches a -1018 file that the database engine refuses to open. Univik Software skips broken sections of such a file. Exchange 2010 databases with a lost folder record send affected mail to Recovered items.
Plan 256 MB beyond your database size in the working folder, because Univik Software duplicates dirty or locked files before opening them.

| Point | Eseutil /p repair | Univik Software |
|---|---|---|
| Live Exchange server needed | Yes | No |
| Source file | Changed by repair | Copy used, source unchanged |
| Preview before export | No | Yes, message by message |
| Output | Mounted database again | MBOX, EML, PST, PDF, MSG, HTML or CSV index |
| Cost | Free | From $99 for 2 PCs |
| Trial | Free | 10 messages per folder |
Export before you repair and a failed hard repair cannot take mail with it. How to Convert EDB to PST lists the export steps.
- Exchange Jet error -1018 means a database page failed its checksum.
- Microsoft puts 99 percent of Exchange 2007 page number mismatch damage on driver, firmware or hardware issues.
- Eseutil /k checks page integrity on a dismounted database.
- Healthy DAG copies patch bad pages. Without a copy, move mailboxes or restore a backup.
- Export mail first, since repair invalidates previous backups.
Univik Software opens a database that logs -1018 so you browse its mailboxes ahead of an eseutil /p repair. Trial edition saves 10 messages per folder. Licences start at $99 for 2 PCs.
Free Download See all featuresQuestions About Exchange Jet Error -1018
Is Jet error -1018 a hardware problem?
Yes. Microsoft's Exchange 2007 guidance puts page level damage behind a page number mismatch on driver, firmware or hardware issues 99 percent of the time, per The database page read failed verification because of a -1018 error (page number mismatch). For Exchange 2010, Microsoft adds that a RAM or firmware fault can raise the error while the page stays undamaged.
Why does my Exchange backup fail with -1018?
Microsoft lists a failed online backup as a symptom of ESE event 474 in Exchange 2010. Backups end with that checksum error. Check storage first and then fix the damaged page.
Can a DAG fix -1018 on its own?
Yes, when a healthy copy exists. In Exchange 2010, Microsoft says page patching replaces corrupted pages with healthy copies, per Understanding Database Maintenance. Lagged copies cannot be patched and need a reseed, per Manage mailbox database copies.
Does eseutil /k fix the bad pages?
No. Exchange 2007 checksum mode verifies database pages only, as Eseutil /K Checksum Mode explains. Fixing needs a page patch, a mailbox move, a restore or a hard repair.
Can I keep using a database with -1018 errors?
No. Microsoft calls -1018 an early warning and tells you to plan for imminent failure or more random damage until you find the root cause, per ESE 474 -1018: Unrecoverable Error Detected in Database. Move mailboxes off the database and fix storage.
How do -1018 and -1019 differ?
Error -1018 is a checksum or page number mismatch. Error -1019 is a blank page where data should sit. Microsoft says online backup does not detect -1019, so it can stay hidden longer.
Does antivirus cause Jet error -1018?
Yes. Exchange 2013 file level scanners lock or quarantine log and database files while Exchange uses them. Microsoft says that can cause -1018 errors in the event log. Anti-Virus Software in the Operating System on Exchange Servers puts .chk, .edb and .log extensions plus the database folders on the exclusion list.
Does Univik Software open a database that reports -1018?
Yes. Univik Software works from an offline duplicate on Windows and scans raw data if the database engine rejects your -1018 file. Trial edition saves 10 messages per folder.