MySQL Optimization Dillema

In response to this post, I have compiled some charts showing the memory usage of my VPS – specifically relating to MySQL, to determine whether or not I should not use InnoDB. Note that, this site is running on an InnoDB schema and so there would be additional migration steps to reap the benefits of disabling this feature. Check out the figures 

Typical usage scenario

total used free shared buffers cached
Mem: 256 252 3 0 0 32
-/+ buffers/cache: 220 35
Swap: 512 96 415

MySQL Running – InnoDB off

total used free shared buffers cached
Mem: 256 110 145 0 0 39
-/+ buffers/cache: 70 185
Swap: 512 34 477

MySQL Stopped

total used free shared buffers cached
Mem: 256 104 151 0 0 40
-/+ buffers/cache: 64 191
Swap: 512 34 477

MySQL Running – InnoDB on

total used free shared buffers cached
Mem: 256 125 130 0 0 41
-/+ buffers/cache: 84 171
Swap: 512 34 477

Apache/MySQL Stopped

total used free shared buffers cached
Mem: 256 42 213 0 0 27
-/+ buffers/cache: 15 240
Swap: 512 26 485

Start Apache

total used free shared buffers cached
Mem: 256 54 201 0 0 27
-/+ buffers/cache: 27 228
Swap: 512 26 485

Start MySQL

total used free shared buffers cached
Mem: 256 83 172 0 0 35
-/+ buffers/cache: 47 208
Swap: 512 26 485

 

In conclusion, I have decided to leave everything as is. It really isn’t worth my time to save ~15MB in memory usage.

Leave a Reply