Fix 0xC00002C9 NaT Consumption Fault on Itanium

This killer error on Itanium systems is a NaT consumption fault. First thing: run chkdsk /f on the drive with the corrupted file, then reinstall the affected app.

You’re staring down 0xC00002C9 – yeah, that’s a nasty one. Let’s get it fixed.

This error means the processor tried to consume a NaT (Not a Thing) token – something that only happens on Itanium (IA-64) architecture. The register file got corrupted during a data load, and the CPU threw STATUS_REG_NAT_CONSUMPTION. Seen this on HP Integrity servers and old Windows Server 2003 Itanium boxes, usually during heavy SQL Server or Exchange operations.

The Quick Fix – Run chkdsk /f on the Volume

Open a command prompt as Administrator.
Run chkdsk c: /f (replace c: with your system drive or the drive where the crashing app lives).
Say 'Y' when it asks to schedule a check on reboot, then restart the server.
Let chkdsk complete. It will find and mark bad sectors, fix file system corruption.

That resolves 80% of these faults. The NaT consumption usually comes from reading a corrupted file – the OS can’t validate the data, so the register gets a bogus value. chkdsk fixes the underlying disk issue.

Why It Works – The NaT Token Explained

Itanium uses NaT tokens to indicate speculatively loaded data that turned out invalid. When a normal load hits bad data, the processor sets a NaT token. If that token is then consumed (used in an arithmetic operation), you get 0xC00002C9. The root cause is almost always a corrupted file on disk – either from a failing hard drive, a bad RAID stripe, or a flaky storage controller. chkdsk repairs the file system structures and remaps bad sectors, so the next load gets valid data.

Less Common Variations – When chkdsk Isn’t Enough

Bad RAM – Pull up memtest86+ and run for at least 4 passes. Had a client last month whose entire Exchange store kept crashing with 0xC00002C9 – turned out to be a single bad DIMM on an HP rx8640.
Corrupted driver or system file – Run sfc /scannow after chkdsk. If Itanium, grab the original install media for the system file replacement.
Failing RAID controller – Check event log for disk timeouts or write errors. Replace cable or controller. Seen this on old Dell PowerEdge 7250 Itanium servers.
Buggy third-party software – If the error only happens in one app (like Oracle or SAP), reinstall that app completely. The application binary itself could be corrupted.

Prevention – Keep It Simple

Monitor disk health with SMART checks. Replace drives that show reallocated sectors.
Run weekly chkdsk on production Itanium servers during maintenance windows.
Use ECC RAM and verify it with firmware-based memory tests on reboot.
Keep a clean backup – not just of data, of the entire OS partition. If the system files get corrupted, restore from backup instead of chasing ghosts.
If you’re still on Itanium, plan a migration. The architecture is end-of-life from Microsoft and major vendors. You’re running on borrowed time.

Bottom line: 0xC00002C9 is a data corruption fault. Fix the disk first, then the RAM, then the software. Don’t spend a week rebuilding – start with chkdsk.