P
P
pixik2018-01-18 21:57:30
Oracle
pixik, 2018-01-18 21:57:30

How to limit memory consumption in Oracle db 11g EE?

Good time!
On the test server (linux) there are almost 4 gigabytes of RAM, when starting the database (Oracle Database 11g Enterprise Edition Release 11.2.0.4.0) 1.2G is used. After the queue of requests (10-20k), about 40 MB of free RAM remains and the server stops responding. Tell me where to dig, what logs to see, what to tweak?
Is it possible to somehow make the database respond more slowly, but not eat up all the possible memory on the server?

Answer the question

In order to leave comments, you need to log in

3 answer(s)
E
Eugene, 2018-01-18
@Lorien_Elf

When starting in the alert log, the database writes non-default parameters. Post them here. First of all, the values ​​of sga_target, pga_aggregate_target, memory_target, sga_max_size, memory_max_size are of interest.
The alert log is somewhere in /u01/app/oracle/diag/rdbms/db_name/db_name/trace/alrert_db-name.log

I
Ivan Sorokin, 2018-01-19
@ivandest

This is for Oracle 9, but I think the difference is not big
RAM=4G: (if one instance is running on the server)
SHARED_POOL_SIZE=512M
DB_CACHE_SIZE=2G
LARGE_POOL_SIZE=16M
PGA_AGGREGATE_TARGET=128M
You can view the parameter values ​​like this:

select name, value/1024/1024||' Mb' from v$parameter where name in ('shared_pool_size', 'db_cache_size', 'large_pool_size', 'pga_aggregate_target');

A
andreev_aa, 2018-02-21
@andreev_aa

If after starting the database, the memory remains, and disappears after the start of work, then either user sessions eat it, then you need to look in the direction of reducing the pga_aggregate_target parameter. Also, with such a small amount of memory, the server can begin to actively use the swap partition. In any case, you need a database alert log for analysis.

Didn't find what you were looking for?

Ask your question

Ask a Question

731 491 924 answers to any question