-->-----Original Message----- -->From: Jeremy Zawodny [mailto:[EMAIL PROTECTED] -->Sent: Tuesday, September 30, 2003 11:24 AM -->To: Dathan Vance Pattishall -->Cc: [EMAIL PROTECTED] -->Subject: Re: What are the effects of key_buffer on a dedicated slave -->[also] --> -->On Tue, Sep 30, 2003 at 11:07:59AM -0700, Dathan Vance Pattishall wrote: -->> -->> -->-----Original Message----- -->> -->From: Jeremy Zawodny [mailto:[EMAIL PROTECTED] -->> -->Sent: Tuesday, September 30, 2003 10:23 AM -->> -->To: Dathan Vance Pattishall -->> -->Cc: [EMAIL PROTECTED] -->> -->Subject: Re: What are the effects of key_buffer on a dedicated slave -->> -->> -->> -->Is that all your slave is doing? Replicating from the master? Are -->> there -->> -->no other queries being run against it? -->> --> -->> -->> It takes 50% of all reads from the application the db supports. --> -->The first thing I'd do is figure how how well utilized the key buffer -->is today. Either grab a copy of mytop (it does it for you), or look -->at SHOW STATUS to compute the percentage based on the ratio of -->key_read_requests to key_reads.
Yes, I use a custom mytop (sent my patches in to you). In fact I'm making a signed java applet to simulate mytop, just to be fancy ;) as well as not having to ssh into a central box that can reach all my servers. --> -->If you're already hitting the buffer 99% of the time, there's little -->point in increasing the size. I figured that but does it really matter on a dedicated slave handling some reads, since there can never be parallel writes? I may be missing the concept. My Goal is to speed up replication (3.23.57) so I can upgrade to 4.0.15 once the master / slave is in-sync and find out if the var key_buffer on a dedicated slave helps out significantly. --> -->Jeremy -->-- -->Jeremy D. Zawodny | Perl, Web, MySQL, Linux Magazine, Yahoo! --><[EMAIL PROTECTED]> | http://jeremy.zawodny.com/ --> -->MySQL 4.0.15-Yahoo-SMP: up 16 days, processed 586,536,970 queries -->(405/sec. avg) -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
