|
|
| (8 intermediate revisions by the same user not shown) |
| Line 3: |
Line 3: |
| == MySQL == | | == MySQL == |
|
| |
|
| === Performance Tuning ===
| | [[MySQL#Performance_Tuning | MySQL Performance Tuning]] |
|
| |
|
| '''1. First, Check innodb and myisam settings'''
| |
|
| |
|
| Run the below to get the size of data in INNODB and MYISAM tables:<br />
| | == OOM Killer == |
| <pre>mysql -e 'SELECT SQL_NO_CACHE Engine, count(*) "# of Tables", concat(round(sum(table_rows)/1000,2),"K") "Rows", concat(round(sum(data_length)/(1024*1024),2),"M") "Size of Data", concat(round(sum(index_length)/(1024*1024),2),"M") "Size of Indexes", concat(round(sum(data_length+index_length)/(1024*1024),2),"M") "Total Size", concat(round(sum(index_length)/sum(data_length)*(100),2), "%") "% Data in Indexes" FROM information_schema.TABLES GROUP BY engine ORDER BY sum(data_length+index_length) DESC LIMIT 10;'</pre>
| |
|
| |
|
| In /etc/my.cnf, change:<br />
| | OOM Killer kills the process with the highest memory usage to prevent the system from running out of memory. It often triggers due to memory load. |
|
| |
|
| <code>innodb_buffer_pool_size</code> - set this to the value from the Total column, innodb row. (Note: if server has lots of RAM, and lots of data, set the innodb_buffer_pool_size = 1G, then set innodb_buffer_pool_instances = 4 (or smaller, or larger, depending on RAM))
| | === Prevent OOM Killer from killing MySQL === |
|
| |
|
| <code>key_buffer_size</code> - set this to the value from the Index column, myisam row. | | You can create a cron job to keep the PID for MySQL updated in the OOM Killer exclusion list: |
| | <code>*/1 * * * * root pgrep -f "/usr/bin/mysql" | while read PID; do echo -1000 > /proc/$PID/oom_score_adj; done</code> |
|
| |
|
| | == References == |
|
| |
|
| === MySQL Error log ===
| | <ul> |
| | | <li>[https://backdrift.org/oom-killer-how-to-create-oom-exclusions-in-linux Backdrift.org - OOM Killer exclusions]</li> |
| To check the MySQL error log, run:<br />
| | </ul> |
| <code>mysql -e "show variables like '%log_error%';"</code> | |
| | |
| This will show the location of the MySQL error log, which you can then grep, tail, etc. (You're looking for stuff like innodb errors, disk space issues, connection messages, etc)
| |
| | |
| | |
| === View MySQL Variables ===
| |
| | |
| You can view the variables that MySQL is using with:
| |
| | |
| <code>mysql -e "show variables like 'max_connections%';" </code> - This shows the max_connections variable. You can substitute any MySQL variable here.
| |
| | |
| This is helpful for database connection errors, or if a change to /etc/my.cnf isn't 'sticking' (In that case, look for other my.cnf files)
| |
| | |
| | |
| === View Currently Running Queries ===
| |
| | |
| <code>mysqladmin process stat</code> - This will show the queries that are running, how many threads, Open tables, slow queries, and a bunch of other helpful stuff. (Note that you can "tail -f" by running: <code>watch mysqladmin process stat</code> )
| |