Xref: alice comp.databases.oracle.misc:41998 comp.databases.oracle.server:67387 Path: alice!news-feed.fnsi.net!hammer.uoregon.edu!logbridge.uoregon.edu!newsfeed.stanford.edu!paloalto-snf1.gtei.net!news.gtei.net!inet16.us.oracle.com!not-for-mail From: Thomas Kyte Newsgroups: comp.databases.oracle.misc,comp.databases.oracle.server Subject: Re: log_buffer settings Date: Sun, 26 Sep 1999 14:00:04 -0400 Organization: Oracle Service Industries Lines: 55 Message-ID: <5l3uNxd7fFLf02fzPpHtKmWtNBkT@4ax.com> References: <7slfc7$bcb$1@nntp3.atl.mindspring.net> Reply-To: tkyte@us.oracle.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: inet16.us.oracle.com 938368744 3942 130.35.34.237 (26 Sep 1999 17:59:04 GMT) X-Complaints-To: usenet@inet16.us.oracle.com NNTP-Posting-Date: 26 Sep 1999 17:59:04 GMT X-Newsreader: Forte Agent 1.6/32.525 A copy of this was sent to "Steve Perry" (if that email address didn't require changing) On Sun, 26 Sep 1999 10:52:45 -0700, you wrote: >Can anyone help out with "realistic" settings for the log_buffer parm, as >well as anything else I might want to check? >I have kept increasing the log_buffer parm to 16meg now, but still keep >seeing the waits below. I changed the log_simultaneous_copies to 16 (2 x >#cpu's), but nothing seems to help. The below was from a 2 hour window when >normal activity was taking place. It's an OLTP system (400 concurrent >users), Oracle 7.3.4, aix 4.3.2 (4 gig of ram), sga of about 700 meg, 250 >meg of shared pool. The Oracle manuals state that increasing the log_buffer >beyond 1 meg does no good, but I've heard otherwise. It hasn't made any >difference for me... There must be something I'm missing. > reduce that log buffer size, get it back down into KILOBYTES, not megabytes. increasing the log buffer will not help (and can in fact hurt). the redo log space requests are not waits for space in the redo log buffer - they are waits for space in the online redo logs themselves. (some documentation erroneously tells you to increase the log_buffer -- it is wrong, increasing the log_buffer -- as you have seen -- will have no effect on this). Increase the size and or number of your online redo logs. Look in your alert.ora file for messages to the effect of "cannot allocate new log". They will be either because "checkpoint not complete" (add more log for this) or because the archiver hasn't finished (tune your redo disks to remove contention between lgwr and arch and perhaps add more redo). >from utlstat report (system statistics) > >Statistic Total Per Transaction Per >Logon Per Second >redo log space requests 10,998 .03 >38.73 1.19 >redo log space wait time 1,070,620 3.38 3,769.79 >115.96 > >Thanks, >Steve > > -- See http://govt.us.oracle.com/~tkyte/ for my columns 'Digging-in to Oracle8i'... Current article is "Part I of V, Autonomous Transactions" updated June 21'st Thomas Kyte tkyte@us.oracle.com Oracle Service Industries Reston, VA USA Opinions are mine and do not necessarily reflect those of Oracle Corporation