ORA-600 [kmgss_activate_granule_1] and Instance Crash After Change of DB_xyz_CACHE_SIZE (Doc ID 745189.1)

Last updated on NOVEMBER 28, 2016

Applies to:

Oracle Server - Enterprise Edition - Version: 10.2.0.1 to 10.2.0.4 - Release: 10.2 to 10.2

Symptoms

The MMAN Memory Manager background process raised an ORA-600 internal error [kmgss_activate_granule_1] and crashed the database instance right after attempt to dynamically change one of the following init.ora parameters:

 STREAMS_POOL_SIZE

 DB_KEEP_CACHE_SIZE

 DB_RECYCLE_CACHE_SIZE

 DB_nK_CACHE_SIZE with n in (2, 4, 8, 16, 32)

The generated trace file named <sid>_MMAN_<pid>.trc shows:

- kmgss_activate_granule on the Call Stack

- one of the the following:

 STREAMS_POOL_SIZE
 DB_KEEP_CACHE_SIZE
 DB_RECYCLE_CACHE_SIZE
 DB_2K_CACHE_SIZE
 DB_4K_CACHE_SIZE
 DB_8K_CACHE_SIZE
 DB_16K_CACHE_SIZE
 DB_32K_CACHE_SIZE

listed under the section entitled DYNAMICALLY MODIFIED PARAMETERS written at the bottom of the MMAN trace file.

Changes

We have seen occurrences of this issue after dynamic change of one of the following init.ora parameters:
 STREAMS_POOL_SIZE

 DB_KEEP_CACHE_SIZE

 DB_RECYCLE_CACHE_SIZE

 DB_nK_CACHE_SIZE with n in (2, 4, 8, 16, 32)

Automatic Shared Memory Management is configured.
In other words, SGA_TARGET needs to be set to a non-zero value to potentially encounter this issue.

Cause

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