|
29 Jun |
Profiling with MySQL Computer, MySQL, Tutorials
|
Whenever you encounter a query slower than you expected, and you’re sure you’ve done everything in your power to ensure your tables are properly optimized and indexed, you might want to profile it. MySQL has an inbuild profiler, which allows you to see very detailed for what part of the query how much time has been spend.
To use it, follow these steps:
(1) Activate the profiler.
Since the profiler is deactivated by default we first have to start it.
SET profiling = 1;
(2) Execute your query.
(3) Find out the query id for profiling.
SHOW PROFILES;
It will return you something like this:
Query_ID | Duration | Query
---------+-----------+-----------------------
... | ... | ...
29 | 0.0006200 | SHOW STATUS
30 | 0.3600000 | SELECT (your query here)
... | ... | ...
Now you know the query id is 30.
(4) Profile the query.
SHOW PROFILE FOR QUERY 30; // example
This will return you the details, which might look like this:
Status | Duration
--------------------------------+-------------------
starting | 0.000010
checking query cache for query | 0.000078
Opening tables | 0.000051
System lock | 0.000003
Table lock | 0.000008
init | 0.000036
optimizing | 0.000020
statistics | 0.000013
preparing | 0.000015
Creating tmp table | 0.000028
executing | 0.000602
Copying to tmp table | 0.000176
Sorting result | 0.000043
Sending data | 0.080032
end | 0.000004
removing tmp table | 0.000024
end | 0.000006
query end | 0.000003
freeing items | 0.000148
removing tmp table | 0.000019
closing tables | 0.000005
logging slow query | 0.000003
cleaning up | 0.000004
In this example, most of the time was actually spend sending the data from the server back to the client. Maybe try narrowing down the amount of data you get from the query. Maybe raise your query cache. Maybe do something different – its entirely up to your database, server and needs what to make of this result. But now you know where to look!