Mysql join with group by query optimization

I am using the following query to get the latest articles (article title + author name + total comments) but it takes long time to run:

SELECT article.id, title, username, count( comment.id ) AS total_comments
FROM article
    LEFT JOIN COMMENT ON comment.article_id = article.id
    LEFT JOIN user ON article.user_id = user.id
WHERE STATUS = "open"
    AND section = "mobiles"
GROUP BY article.id
ORDER BY article.id DESC 
LIMIT 0 , 30

The output of explain is:

id  select_type     table   type    possible_keys   key     key_len     ref     rows    Extra
1   SIMPLE  article     ALL     status  NULL    NULL    NULL    77  Using where; Using temporary; Using filesort
1   SIMPLE  comment     ref     article_id  article_id  5   test.article.id     2    
1   SIMPLE  user    eq_ref  PRIMARY     PRIMARY     4   test.article.user_id    1

how to re-write the query to avoid creating temporary table?

Storage engine is MyISAM. Here is tables details wit开发者_运维技巧h indexes:

article 
id, title,  body,   user_id,    status,     section
primary key: id
indexes: (user_i),(status,section,id)


comment
id, article_id, user_id,    body
primary key: id
index:(article_id)


user
id, username
primary key: id

Would using a sub query instead of the grouped join speed up the process.

SELECT article.id, title, username, 
      (
       select count(*) from COMMENT 
       where comment.article_id = article.id 
       ) AS total_comments 
FROM article
    LEFT JOIN user ON article.user_id = user.id 
WHERE STATUS = "open" 
    AND section = "mobiles" 
ORDER BY article.id DESC  
LIMIT 0 , 30 

as per a previous question How to make a nested query?

I hope this helps.

This execution plan is pretty good. You really can't avoid creating and sorting the temp table: your query calls for a grouped and sorted summary, which is, ummm, a temporary table. If you give your mySQL server process more RAM you may be able to get it to use an in-memory table rather than an on-disk table.