I'm upgrading to a faster server with whm/cpanel and the mysql databases will be on a SSD. I've read a bit about not wanting any query cache and stuff like that, but as a novice, I've learned that tinkering with my.cnf can be disastrous.

Here's the current my.cnf for my "hard drive" based webserver.




What would you change in it on if mysql is on a SSD?

Here's a bit of info from tuner.sh all:

Avg. qps = 447

Current thread_cache_size = 256
Current threads_cached = 96
Current threads_per_sec = 0
Historic threads_per_sec = 0
Your thread_cache_size is fine

Current max_connections = 250
Current threads_connected = 20
Historic max_used_connections = 116
The number of used connections is 46% of the configured maximum.
Your max_connections variable seems to be fine.

Current InnoDB index space = 794 M
Current InnoDB data space = 1.10 G
Current InnoDB buffer pool free = 14 %
Current innodb_buffer_pool_size = 1.95 G
Depending on how much space your innodb indexes take up it may be safe
to increase this value to up to 2 / 3 of total system memory

Max Memory Ever Allocated : 3.62 G
Configured Max Per-thread Buffers : 3.02 G
Configured Max Global Buffers : 2.21 G
Configured Max Memory Limit : 5.23 G
Physical Memory : 15.56 G
Max memory limit seem to be within acceptable norms

Current MyISAM index space = 222 M
Current key_buffer_size = 128 M
Key cache miss rate is 1 : 1115
Key buffer free ratio = 72 %
Your key_buffer_size seems to be fine

Query cache is enabled
Current query_cache_size = 128 M
Current query_cache_used = 28 M
Current query_cache_limit = 32 M
Current Query cache Memory fill ratio = 22.42 %
Current query_cache_min_res_unit = 4 K
Your query_cache_size seems to be too high.
Perhaps you can use these resources elsewhere
MySQL won't cache query results that are larger than query_cache_limit in size

Current sort_buffer_size = 4 M
Current read_rnd_buffer_size = 4 M
Sort buffer seems to be fine

Current join_buffer_size = 4.00 M
You have had 166 queries where a join could not use an index properly
join_buffer_size >= 4 M
This is not advised
You should enable "log-queries-not-using-indexes"
Then look for non indexed joins in the slow query log.

Current open_files_limit = 120260 files
The open_files_limit should typically be set to at least 2x-3x
that of table_cache if you have heavy MyISAM usage.
Your open_files_limit value seems to be fine

Current table_open_cache = 60000 tables
Current table_definition_cache = 2000 tables
You have a total of 733 tables
You have 851 open tables.
The table_cache value seems to be fine

Current max_heap_table_size = 2.97 G
Current tmp_table_size = 2.97 G
Of 262494 temp tables, 33% were created on disk
Perhaps you should increase your tmp_table_size and/or max_heap_table_size
to reduce the number of disk-based temporary tables
Note! BLOB and TEXT columns are not allow in memory tables.
If you are using these columns raising these values might not impact your 
ratio of on disk temp tables.

Current read_buffer_size = 128 K
Current table scan ratio = 1005 : 1
read_buffer_size seems to be fine

Current Lock Wait ratio = 1 : 101097
Your table locking seems to be fine

The OLD web server is an 8 core single Xeon @ 16GB ram. The NEW web server is a 24 core dual Xeon @ 64GB ram. The SSD on the NEW server will contain the OS and the Database only, and all the sites are WordPress based resulting in about 250,000 pageviews total a day.

Thanks so much for any help or suggestions!



  • "I've read a bit about not wanting any query cache..." Says who? – ceejayoz Oct 2 '14 at 20:32
  • @ceejayoz, we at Percona frequently see scaling problems that are fixed by disabling the query cache. I advise to disable it unless you see a very favorable ratio of QC hits/misses, perhaps 10:1 or more. – Bill Karwin Oct 2 '14 at 20:40
  • @ceejayoz Query Cache hit rate was 44%, and I disabled query cache, all is well. :) – Ben The Novice Oct 6 '14 at 19:24
  • Percona has a great configuration wizard, i am very satisfied with the result: tools.percona.com/wizard – da_didi Oct 29 '14 at 8:16


It makes little sense to set this lower than table_open_cache. See the default value described in https://dev.mysql.com/doc/refman/5.6/en/server-system-variables.html#sysvar_open_files_limit

table_open_cache=25000 table_cache=60000

You probably don't need to set this so high considering you only have < 900 tables open.

Note that table_cache is an obsolete synonym for table_open_cache. The last one in the my.cnf file takes priority. So you have set your table_open_cache to 60000, overwriting the earlier setting. I suggest deciding on the value you want and use one line to set it, to avoid confusion.


This is fine.


The key buffer is only for MyISAM. If you have no MyISAM tables, the default of 8M is fine.

If you have lots of MyISAM tables, you should convert them to InnoDB. ;-)


Since you upgraded to 64G, this seems low. On the other hand, you have only about 2G of InnoDB data and indexes, which is not unusual for a simple Wordpress site. There's no benefit in increasing the buffer pool any higher.


The default number of BP instances is 8 as of MySQL 5.6.6. Since you have 24 cores, you might want to split the BP at least up to 8 instances.


16x the default of 256K seems too high. Ideally, you shouldn't need to rely on the join buffer because joins should be handled by indexes.

In MySQL 5.6, the join buffer is also used for Batched Key Access optimization, but this doesn't kick in very often.


Also 16x the default of 256K. Useful for avoiding sort merge passes when you do filesorts of large tables. Monitor SHOW GLOBAL STATUS LIKE 'Sort_merge_passes', but I bet you can decrease this.


Also 16x the default of 256K. Used for MyISAM tables, or for Multi-Range Read optimization for any storage engine. Very rare to need this.

tmp_table_size=3048M max_heap_table_size=3048M

These are probably excessive. I start out small, and increase tmp_table_size if I see a steady increase in SHOW GLOBAL STATUS LIKE 'Created_tmp_disk_table'.


Probably unnecessary to have a thread cache at all on your system. The default would be 33. On a modern version of Linux, creating a new thread is quick anyway, so the need for the thread cache is not as important is it is on Solaris, FreeBSD, etc.

max_connections=250 table_cache=60000

Note that table_cache is an obsolete synonym for table_open_cache. The last one in the my.cnf file takes priority. So you have set your table_open_cache to 60000, overwriting the earlier setting. I suggest deciding on the value you want and use one line to set it, to avoid confusion.

  • Thank you Bill Karwin -- I'm honored that you replied, and all of your changes worked perfectly. – Ben The Novice Oct 6 '14 at 19:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy