> Minor correction: The post i point to is about the slow log, but I presume > also relevant for the general log. And the good comments I mentioned come in > the followup posting at http://www.bitbybit.dk/carsten/blog/?p=116 > > / Carsten
Thanks Carsten, I read the comments and Sheeri mentions mysqlsla which I have already tried. Back to square one. I might look at munin again and see if someone has written a plug-in that graphs query type but that seems too much hassle for my situation. I have the raw data and I want the appropriate tool to analyze it. Part of the reason is that the data is from a MyISAM based web app and I am writing a report recommending it be moved to a transactional storage engine. AIUI a rule of thumb is that if between 15% - 20% of statements are non SELECT/INSERT then one can obtain equal or better performance with something like InnoDB. That being said, the benefits of InnoDB (good recovery features, transactions, advanced indexes, foreign key contraints) make it a good default choice and I will recommend it anyway. Plus we have order processing stuff going on and it seems right to have "atomicity" in that process. It would be a bit better though to confidently state that the query-mix skews it towards InnoDB... if I can only prove it :-) -- GPG Key fingerprint = B323 477E F6AB 4181 9C65 F637 BC5F 7FCC 9CC9 CC7F -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/mysql?unsub=arch...@jab.org
