On Fri, Aug 01, 2003 at 01:00:13PM -0400, Jack Coxen wrote: > > Since starting up RTG not quite 3 months ago, my database has grown to over > 10GB in size. If I can get the drive space to keep the 2 years worth of > data I want to, my database will be edging upward toward 100GB. There are > currently 682 tables ranging in size from 2 KB/1 record to around 310 > MB/7570511 records depending on the number of ports on a router and how much > traffic flows through it. > > I'm running around 2500 queries/minute - 99+% of them inserts. My cpu load > is minimal - around .40 nominally but it may go as high as 1.80 or so when > handling multiple large retrievals. If I look at the iostats output for my > server, the drive controller bandwidth utilization is around 30-40% during > normal operation but immediately jumps to 100% utilization during retrieves. > Retrieval is done via Perl scripts or from Perl/PHP WebPages. > > I'm trying to get a RAID array loaded with striped drives to hang off the > server but until I can talk someone into signing the check, I need to do > anything I can to improve performance.
This sounds like it'll be a classic disk I/O problem. Sooner or later you hit the wall on what a single disk is able to do for you. The only real solution is a faster disk or adding more disks to spread the work around. Adding memory will help a very small amount. Storing less data, of course, would too. Making the data take less space will also help. But it sounds like you may have done some of that already. Jeremy -- Jeremy D. Zawodny | Perl, Web, MySQL, Linux Magazine, Yahoo! <[EMAIL PROTECTED]> | http://jeremy.zawodny.com/ MySQL 4.0.13: up 6 days, processed 214,041,264 queries (397/sec. avg) -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
