I've had a look at K2's DB performance and found that there are some optimizations that could be made:Some indexes for better/faster access:ALTER TABLE `jos_k2_comments` ADD INDEX
`pub_id` (`published`, `itemID`);
ALTER TABLE `jos_k2_tags` ADD INDEX
`pub_id` (`published`, `id`);
ALTER TABLE `jos_k2_items` ADD INDEX
`state_access` ( `published` , `trash` , `access` , `publish_up` , `publish_down`
, `catid`, `featured`, `created` DESC);        
In the file /components/com_k2/models/item.php line 779, change$query="SELECT * FROM #__k2_tags as tags WHERE tags.published=1 AND tags.id IN (SELECT tagID FROM #__k2_tags_xref WHERE itemID=".(int)$itemID.")";for $query="SELECT * FROM #__k2_tags as tags JOIN #__k2_tags_xref ON tags.id = tagID AND tags.published=1 AND itemID=".(int)$itemID;

If you create an index to mysql table, does mysql do as some database prgrams do and optomze the query and use the best index if it exists and if it doesnt have the necessary index, stop and create a tempory one to produce the desired results?
i come from vfp and rushmore optimization environments

PHP is a acripting language, so my guess is that DB accesses are treated dynamically, and therefore they choose the best index that already exists, but AFAIK, mysql doesn't create 'temporary indexes'...