Re: Understanding disk sorting..

yong321_at_yahoo.com (Yong Huang) wrote in message news:<b3cb12d6.0411150739.46e4d4af_at_posting.google.com>...
> netcomradeNSPAM_at_bookexchange.net (NetComrade) wrote in message news:<419505fc.332673278_at_localhost>...
> > I have a query that does the following
> > insert into table1
> > as select
> > select id, sum(col1), sum(col2), sum(col3)....
> > where some_id=:var
> > some_other_id=:var2
> > group by id
> >
> > When i look at the trace file I see:
> > WAIT #1: nam='db file scattered read' ela= 2 p1=18 p2=141926 p3=128
> >
> > However, direct path reads and writes are _always_ block by block. Do
> > I just not understand how temp sorts work? Is it the expected
> > behavior? I tried to play with sort_multi_block_read_count
> > .......
> > We use Oracle 8.1.7.4 on Solaris 2.7 boxes
>
> File# 18 is probably the regular datafile that has your selected table in it.
>
> Yong Huang

I suspect netcomrade was asking why a block by block would be db file scattered read, and I suspect the answer is semantic: The scattered read refers to discontinous memory locations to put the data, not what the disk head is doing.

http://download-west.oracle.com/docs/cd/B10501_01/server.920/a96533/instance_tune.htm#15945

jg

--
@home.com is bogus.
http://www.signonsandiego.com/uniontrib/20041113/news_1b13insure.html