Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: *Very* high 'Fetch/elapse' cell in TKPROF report: cause ?

Re: *Very* high 'Fetch/elapse' cell in TKPROF report: cause ?

From: Sybrand Bakker <sybrandb_at_hccnet.nl>
Date: Thu, 26 Aug 2004 15:13:20 +0200
Message-ID: <bcori05s1nih9st8em1lltrr8qtqodp9qt@4ax.com>


On 26 Aug 2004 04:59:05 -0700, spendius_at_muchomail.com (Spendius) wrote:

>Where do these pretty high fetch durations come from ? Where
>should we look ?
>
>Thanks.
>
>
>SQL> show sga
>Total System Global Area 1079217616 bytes
>Fixed Size 732624 bytes
>Variable Size 889192448 bytes
>Database Buffers 188743680 bytes
>Redo Buffers 548864 bytes
>
>Optimizer-related params values (the SAME as on the 8i version
>except for the block size):

The last line in what I quoted basically says it all. A different block size will be a *MAJOR* factor in performance differences, *especially* if all the other parms are identical, *including* the parameters where the unit is a *block*

--
Sybrand Bakker, Senior Oracle DBA
Received on Thu Aug 26 2004 - 08:13:20 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US