How to resolve database crash issues with large memory_target set to over 250 Gig (Doc ID 1677008.1)

Last updated on APRIL 26, 2016

Applies to:

Oracle Database - Enterprise Edition - Version 12.1.0.1 and later
Information in this document applies to any platform.

Goal

This note helps you in resolving database startup issues with MEMORY_TARGET set to over 250 GB 

The alert log can report the following errors:

Shared IO Pool defaulting to 512MB. Trying to get it from Buffer Cache for process 38521.

Shared IO Pool size defaulting to 536870912, 559, 1, 536870912
Special Shared IO pool request 536870912 with Auto-SGA

The database starts up fine with MEMORY_TARGET set to 250GB, and there were no failures after that startup.  The database tries to startup with memory target is at 850GB then all the processes will die, the database becomes inaccessible, and the database can't be restarted without restarting the server.

The instance DOES start with 850GB, but dies after a few minutes and does not restart. 

Followed the recommendations outlined  in Section B.6 of the following reference

Oracle® Database Administrator's Reference 12c Release 1 (12.1) for Linux and UNIX-Based Operating Systems
B Administering Oracle Database on Linux
B.6 Allocating Shared Resources

http://docs.oracle.com/database/121/UNXAR/appc_linux.htm#UNXAR011

To use the MEMORY_TARGET or MEMORY_MAX_TARGET feature, the following kernel parameters must be modified.

/dev/shm mount point should be equal in size or larger than the value of SGA_MAX_SIZE, if set, or should be set to be at least MEMORY_TARGET or MEMORY_MAX_TARGET, whichever is larger. For example, with MEMORY_MAX_TARGET=4GB only set, to create a 4 GB system on the /dev/shm mount point:

Run the following command as the root user:

# mount -t tmpfs shmfs -o size=4g /dev/shm

Ensure that the in-memory file system is mounted when the system restarts, add an entry in the /etc/fstab file similar to the following:

tmpfs /dev/shm tmpfs size=4g 0

 

The /etc/fstab should have the following entries similar to the following to accommodate shared memory of over 850Gig:

shmfs                   /dev/shm                tmpfs   size=855g       0 0

This is discussed in the documentation 

Solution

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