>>Jonathan Hilgeman wrote: >> >>>Currently, I have a Celeron 1.2 Ghz server with 512 RAM, and I'm >>>considering moving to a P4 2 Ghz with the same amount of RAM. I have >>>a few specific tables with several million rows of data, and it >>>takes quite a long time to process that data on my current server. >>>Does anyone have a good idea of the type of performance increase I'd >>>see if I moved to a P4 server? >> >>It's going to depend of in your queries are disk bound or processor >>bound. Check the processor usage when a big query is running (and >>not much else is running). >> >>If the CPU usage is at 80-100% the query is processor bound, so >>you'll see some improvement (though I can't tell you how much, other >>than the clock rate scaling). >> >>On the other hand, if the CPU usage is low, you are likely disk >>bound, so a faster processor won't help much. It would be better to >>add RAM, tune your queries or get faster disks (or all three) in this >>case.
[EMAIL PROTECTED] wrote: >scuse the ignorance are u saying high cpu usage is better used on one >query or better be freed up for other queries ? ram is always an issue >as the queries get buffered I'm saying that if your current machine has the mysqld CPU utilization at 100% then what's keeping your queries from going faster is that you've run out of CPU capacity, so adding more (via a faster processor) will make the query go faster. If the mysqld CPU utilization is at 25% then the speed of the processor is not what's making the query go slow (you still have processor capacity left to use, but you can't utilize it because something else is slowing you down). In this case a faster processor won't help much...the processor isn't the limiting factor. This assumes you're doing the benchmarking with nothing else running, obviously. If MySQL is at 25% but you have other programs running heavily it doesn't tell you much, but if it's at 25% and the rest of the capacity is idle then the CPU isn't the problem. --Ware Adams -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
