Hi Terence, key_buffer_size is the amount of memory that is used to cache your index tables. If you are using 100% of it most of the time, you may benefit from an increase in it if you can afford to. You can also look at show status and look at key_reads and key_writes to give you an idea of how often you are using your index cache. Basically, the more indexing you can store in memory the faster your queries tend to be (if they are structured to take advantage of the indexes).
You can get more info by looking at: http://www.mysql.com/doc/en/Server_system_variables.html When MySQL starts up, the cache is empty, and is loaded upon use. One cool new feature in 4.1.1 onwards is the ability to pre-load the cache, and to create seperate index caches for use by different tables. Hope this helps. Eric "Terence" <[EMAIL PROTECTED]> wrote in message news:[EMAIL PROTECTED] > Hi, > > Can somebody tell me whether this is good or bad: > Our mysql server has > key_buffer_size = 402,653,184 (without the commas) > > The MySQL administrator tool indicates a 100% usage most of the time, and > 362,324,992 as current usage within 24 hours which builds up very fast. Are > there memory leaks for threads on my system? Or is this normal behaviour? > Restarting the mysql server clears the above. I can't seem to find much > information on the mysql site or documentation as to what the ideals > are. > > Running MySQL 4.1.1 > Dual Xeon 2.4GhZ > 1 Gig Ram > Redhat 9 > > Thanks > > > -- > MySQL General Mailing List > For list archives: http://lists.mysql.com/mysql > To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED] > > -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
