My Oracle Support Banner

Data Corruption fixes in Red Hat AS 2.1 e.24 kernel (Doc ID 241820.1)

Last updated on APRIL 16, 2020

Applies to:

Linux OS - Version 2.4.9 and later
Oracle Cloud Infrastructure - Version N/A and later
Red Hat Advanced Server
Linux x86


Fixes for data corruption in Red Hat AS 2.1 e.24 kernel related cciss drivers.

Two potential data corruption scenarios have been identified. These scenarios can occur under heavy, complex I/O loads.

The first scenario only occurs while performing memory mapped file I/O, where the file is simultaneously unlinked and the corresponding file blocks reallocated. Furthermore, the memory mapped must be to a partial page at the end of a file on an ext3 file system. As such, Red Hat considers this scenario unlikely.

The second scenario was exhibited in systems with more than 4 GB of memory with a storage controller capable of block device DMA above 4GB (64-bit DMA). By restricting storage drivers to 32-bit DMA, the problem was resolved. Prior to this errata, the SCSI subsystem was
already restricted to 32-bit DMA; this errata extends the restriction to block drivers as well. The change consists of disabling 64-bit DMA in the cciss driver (the HP SA5xxx and SA6xxx RAID controllers). The performance implications of this change to the cciss driver are minimal.

In this scenario you can see from Oracle point of view following problems:
1.- some oracle process reports corruptions like ORA-1578

2 - On the alert log file you can see messages like:
Fractured block found during backing up datafile
Fractured block found during reading block


Possible installation and use of cciss drivers.


To view full details, sign in with your My Oracle Support account.

Don't have a My Oracle Support account? Click to get started!

In this Document

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.