从MySQL检索数据缓慢

Question

I want to retrieve some data from MySQL based on my complicated queries written in my PHP codes. The speed of retrieving data is so slow. How can I solve this problem? Is there any way to improve the performance? I'm using WAMP Server. I need to do it all in PHP, Indeed I want to concatenate all tweets from the same user and push it in an array with his/her name. I have used $twitter array to store the name of twittersand I concatenate all tweets from the same user in dataset like this :

$number_of_twitters=sizeof($twitters);

for($i=0;$i $number_of_twitters;$i++){
    $getTweetsQuery="SELECT Tweets FROM tweettable WHERE Name='$twitters[$i]'";
    $tweets=mysql_query($getTweetsQuery);

$tweetString="";

while($tweetsRow=mysql_fetch_array($tweets)){
    $tweetString .= $tweetsRow[0]." ";
}

So why it is so heavy to be performed ?

Answer 1

Capture the SQL being issued by your application. If you can't do it in the app, you could enable the MySQL general log, but that will log ALL sql statements from all sessions.

Use EXPLAIN to see the generated access plans.

The knee-jerk quick answer is "add indexes", but the key is determining the most appropriate indexes for the queries you are running, and eliminating indexes that are unnecessary.

There may be predicates in your queries that are preventing optimum performance, for example, wrapping columns in functions disables MySQL from performing index range scans and index seeks.

The other knee-jerk quick answer is to add memory to the instance, so that more blocks are cached, to reduce physical I/O.

The MyISAM storage engine takes table level locks, even for SELECT statements, and that kills concurrency. InnoDB performs much better with concurrent queries.

---

UPDATE

The most appropriate index for this query:

SELECT Tweets FROM tweettable WHERE Name='$twitters[$i]'

would be a covering index:

... ON tweetable (Name, Tweets)

An EXPLAIN will show the current execution plan:

EXPLAIN SELECT Tweets FROM tweettable WHERE Name='foo'

Ideally, we'd like the explain output to show 'Using index'. But at a minimum, we want to see MySQL accessing the rows by an index, so it doesn't have to do a full scan of every row in the table to see if the row satisfies the predicate.

It also appears that this query is in a loop. It is usually more efficient to retrieve a set of rows with a single statement, rather than running multiple statements. (And this is particularly the case when each execution is pulling a small number of rows from a very large table, and the query is doing a full scan of the table.)

SELECT t.Tweets
  FROM tweetable t
 WHERE t.Name = 'fee'
    OR t.Name = 'fi'
    OR t.Name = 'fo'
    OR t.Name = 'fum'
 ORDER BY t.Name

or the equivalent:

SELECT t.Tweets
  FROM tweetable t
 WHERE t.Name IN ('fee','fi','fo','fum')
 ORDER BY t.Name

MySQL also has a GROUP_CONCAT function you might find useful.

SELECT GROUP_CONCAT(t.Tweets ORDER BY t.Name SEPARATOR ' ') AS tweetstring
  FROM tweetable t
 WHERE t.Name IN ('fee','fi','fo','fum')

NOTE The length of the string returned by GROUP_CONCAT is limited by the group_concat_max_len and max_allowed_packet variables.