Innodb buffer pool size config is not taking effect

Team,

Have a Single Server that has both App and DB. It has 16 GB Ram. So I tried to increase the innodb_buffer_pool_size to 8GB to improve the DB Performance.

in the /etc/mysql/my.cnf i
image

After that i restarted mysql service.
To Verify, whether the buffer pool is increased, i gave the following command

The Value that is set is 768M and not 8GB. Also the default value given in the my.cnf file is 256M. So not sure how come 768M is selected as the buffer size.

Are we missing some thing here? How do we increase the pool size and verify whether it has actually increase or not?

Thanks,
Saravana
DigixrTechnologies

1 Like

It may help to review the innodb docs and look at the logs, possibly increasing verbosity.

https://mariadb.com/kb/en/innodb/
https://mariadb.com/kb/en/innodb-buffer-pool/
https://mariadb.com/kb/en/error-log/

Sharing the versions of Frappe, ERPNext , operating environment and MariaDB may also help.

finally, we were able to make the config work. We had to update the configuration in the following file /etc/mysql/mariadb.conf.d/erpnext.cnf

Updating the values in the my.cnf is getting overwritten.

1 Like

hi @saravana i am facing the same issue,
my erpnext.cnf looks like this.

# BEGIN ANSIBLE MANAGED BLOCK
[mysqld]
pid-file        = /var/run/mysqld/mysqld.pid
socket          = /var/run/mysqld/mysqld.sock
# END ANSIBLE MANAGED BLOCK

can u confirm the exact format to add poolsize and isntances lines?

Hi. My ERPNext cnf file also looks similar. All configuration are in the other file. @saravana can you share the changes here, I am also running on 16 GB, 6 core Ubuntu 20 server. Getting not so great performance. Thanks.

i used like below and its working.

# BEGIN ANSIBLE MANAGED BLOCK
[mysqld]
pid-file        = /var/run/mysqld/mysqld.pid
socket          = /var/run/mysqld/mysqld.sock
# END ANSIBLE MANAGED BLOCK

innodb_buffer_pool_size	= 32G
innodb_buffer_pool_instances = 32
1 Like

Thanks. Will try it out.