Exchange checksum recovery
Exchange reports JET_errReadVerifyFailure (-1018) when a database page fails checksum or page-number verification after being read. The immediate problem may be one page, but the incident can also indicate a storage-path fault affecting more of the EDB or its log stream.
Understand the symptom
A page carries identity and checksum information that Exchange validates after I/O. Recovery distinguishes isolated page damage from wider media, controller, cache, replication, or misdirected-write problems, then evaluates healthy database copies, backups, logs, and mailbox-level extraction before lossy repair.
What you may see
The failing page did not pass Exchange checksum or page-number validation.
Required pages or recovery operations encounter the damaged block.
A consistency scan reads corruption that normal mailbox activity had not reached.
The affected page may belong to a limited B-tree or message range.
Protect the source
Eseutil /p restores structural consistency by removing unrecoverable pages or records.
The log chain may contain changes needed for clean recovery or a database copy.
Replication actions can replace evidence before copies are compared.
Do this now
Stop repeated mount, repair, reseed, backup, and maintenance attempts where operationally safe.
Save event IDs, page number, database GUID, log generation, copy status, and storage alerts.
Copy EDB, checkpoint, all available logs, backups, configuration, and healthy replica evidence.
Snapshot or image the volume and underlying array when hardware or I/O errors are present.
NEPFOSYS process
NEPFOSYS maps the failed page and compares available database copies, backups, and log continuity.
Replay, restore, extraction, and repair options are tested on copies with loss boundaries documented.
Priority mailboxes, folders, messages, and attachments are checked in the recovered output.
A clear next step
Start a case with the exact symptoms, incident timeline, prior attempts, encryption details, and the files that matter most. NEPFOSYS will route the case to the appropriate DFW intake or nationwide mail-in path.
Start with a case reviewQuestions, answered
A database page read from storage failed checksum or page-number verification.
It may make a database structurally mountable by discarding damaged content. Healthy copy, restore, or extraction paths should be evaluated first.
No. It can involve disks, arrays, controllers, cabling, firmware, cache, virtualization, or earlier writes. Storage must be investigated end to end.
DFW local · Nationwide mail-in