SELECT with WHERE IN and subquery extremely slow

Question

I want to exectute the following query:

SELECT *
FROM `bm_tracking`
WHERE `oid` IN
    (SELECT `oid`
    FROM `bm_tracking`
    GROUP BY `oid` HAVING COUNT(*) >1)

The subquery:

SELECT  `oid` 
FROM  `bm_tracking` 
GROUP BY  `oid` 
HAVING COUNT( * ) >1

executes in 0.0525 secs

The whole query "stucks" (still processing after 3 minutes...). Column oid is indexed.

Table bm_tracking contains around 64k rows.

What could be the reason for this "stuck"?

[Edit: Upon request]

CREATE TABLE `bm_tracking` (
 `id` int(11) NOT NULL AUTO_INCREMENT,
 `oid` varchar(10) NOT NULL,
 `trk_main` varchar(50) NOT NULL,
 `tracking` varchar(50) NOT NULL,
 `label` text NOT NULL,
 `void` int(11) NOT NULL DEFAULT '0',
 `created` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP,
 PRIMARY KEY (`id`),
 KEY `oid` (`oid`),
 KEY `trk_main` (`trk_main`),
 KEY `tracking` (`tracking`),
 KEY `created` (`created`)
) ENGINE=MyISAM AUTO_INCREMENT=63331 DEFAULT CHARSET=latin1

[Execution Plan]

Answer 1

Generally exists EXISTS faster than IN so you can try this and see if it executes better for you

SELECT *
FROM `bm_tracking` bt
WHERE EXISTS
(   SELECT 1
    FROM `bm_tracking` bt1
    WHERE bt.oid = bt1.oid
    GROUP BY `oid` 
    HAVING COUNT(*) >1
)

EDIT:

if you notice from the EXPLAIN you posted... the IN() is considered as a DEPENDENT SUBQUERY which is a correlated subquery... meaning that for every row in the table all rows in the table are pulled and compared... so for example 1000 rows in the table would mean 1000 * 1000 = 1 million comparisons -- thats why its taking such a long time