Chris wrote: [snip] > will not hit the first cache because the fields are different. > http://dev.mysql.com/doc/refman/4.1/en/query-cache-how.html All very true, but when a busy website with 100 hits/interval has 25 duplicate queries, that 24% saved per interval. My only point was - when you're thinking of caching, start with what is readily available instead of re-inventing the wheel elsewhere. >> The number of connections is presumably only important if we speak >> about the number of concurrent connections. If each query can be >> dealt with faster due to caching, the number of concurrent >> connections should drop. > > Err no - it still has to connect to the database (thus use a > connection slot) to even get to the cache. Err yes - a query replied to out of cache will take less time to complete, therefore the connection will be given up faster, therefore less _concurrent_ connections. /Per Jessen, Zürich -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
