On Windows, High Memory Usage Is Seen, And Is Associated With Processes Failures, And Core Dumps.
Last updated on NOVEMBER 28, 2016
Applies to:Oracle Database - Enterprise Edition - Version 10.2.0.1 to 10.2.0.3 [Release 10.2]
Microsoft Windows x64 (64-bit)
***Checked for relevance 19-Nov-2014***
On a Windows 64-bit system, a long running process may fail intermittently with the following set of errors reported to the alert log :
ORA-07445: exception encountered: core dump [ACCESS_VIOLATION] [clsdcxini+62][PC:0x6515858][ADDR:0x18] [UNABLE_TO_READ] 
ORA-07445: exception encountered: core dump [ACCESS_VIOLATION][_intel_fast_memset.A+44][PC:0x6544028] [ADDR:0x0] [UNABLE_TO_WRITE] 
ORA-07445: exception encountered: core dump [ACCESS_VIOLATION] [unable_to_trans_pc][PC:0x6515858][ADDR:0x18] [UNABLE_TO_READ] 
You may also see the below, usually associated with database memory errors:
ORA-04030: out of process memory when trying to allocate xxxx bytes (xxxxxx,xxxxxx)
After the failure, running the same long running process completes without problems but, while monitoring the execution of the long running process, virtual memory consumption is observed to be quite higher than usual.
Please consider the issue being discussed here when having any of the following conditions:
o Oracle 10.2.0.3 or before running on Windows 64-bit.
o ORA-07445 errors generated with the arguments [UNABLE_TO_WRITE] and [UNABLE_TO_READ].
o Low virtual memory seen when running long running processes, that could eventually lead to the failure of the process.
o Database memory errors. i.e. ORA-04030.
No changes have been made either at the database level and the operating system level.
Sign In with your My Oracle Support account
Don't have a My Oracle Support account? Click to get started
My Oracle Support provides customers with access to over a
Million Knowledge Articles and hundreds of Community platforms