| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
 |
 | |||
| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
|
| |||
Home -> Community -> Usenet -> c.d.o.server -> Re: Oracle 9i Performance on SPARCE 450 Solaris 5.7
Thanx! 4 the reply. As per statspack the top waits were...Any good
places to look for....
Instance Efficiency Percentages (Target 100%)
Buffer Nowait %: 99.99 Redo NoWait %: 100.00
Buffer Hit %: 59.37 In-memory Sort %: 99.83
Library Hit %: 91.70 Soft Parse %: 67.68
Execute to Parse %: 61.25 Latch Hit %: 100.00
Parse CPU to Parse Elapsd %: 85.40 % Non-Parse CPU: 95.44
Shared Pool Statistics Begin End
------ ------
Memory Usage %: 27.56 91.76
% SQL with executions>1: 27.56 50.60
% Memory for SQL w/exec>1: 26.81 21.83
Top 5 Wait Events
~~~~~~~~~~~~~~~~~ Wait % Total Event Waits Time (s)Wt Time
-------------------------------------------- ------------ -----------
-------
db file scattered read 8,085,933 3,747
17.04
control file parallel write 135,512 3,594
16.34
log file parallel write 310,126 3,336
15.17
log file sync 293,325 3,189
14.50
async disk IO 1,942,339 2,643
12.02
-------------------------------------------------------------
==============================================================================
"Howard J. Rogers" <howardjr2000_at_yahoo.com.au> wrote in message news:<3PEx9.69357$g9.196611_at_newsfeeds.bigpond.com>...
> You shouldn't particularly be concerned about it at all. The hit ratio is
> largely meaningless.
>
> What did Statspack say about your top 5 wait events? That's more where I'd
> start looking.
>
> Regards
> HJR
>
> "danths" <d_sameer_at_yahoo.com> wrote in message
> news:78e940e1.0211041629.5890223f_at_posting.google.com...
> > Hi All,
> > After having upgraded from 8i to 9i I have been running into a lot
> > of perf issues. For one, the cpu ( 1 processor / 1G mem) has been
> > spiking up a lot.
> > Here are few details. In 8i I always has buffer_cache_hit_ration > 90%
> > , but not anymore. Nothing significant has changed including disk
> > layout. I am already working with Oracle support for the past one
> > week and they haven't been able to suggest anything . I have 1G mem
> > and Oracle is already consuming 1/2 of it and so I cannot increase any
> > SGA and hence db_buffer_cache. I also ran statspack and nothing major
> > had shown up. There's been no swapping either. How concerned should I
> > be regarding the buffer ratio. This is for an OLTP application. Pls
> > suggest.
> >
> > SQL> show sga
> >
> > Total System Global Area 537691892 bytes
> > Fixed Size 280308 bytes
> > Variable Size 419430400 bytes
> > Database Buffers 117440512 bytes
> > Redo Buffers 540672 bytes
> >
> >
> > ~Database Version : Oracle9i Enterprise Edition Release
> > 9.0.1.4.0
> > ~Database Up Since : 02:27:01 AM, October 31 2002
> > !Buffer Cache Hit Ratio : 58.2289
> > ~Library Cache Miss Ratio : 0.1588
> > ~Dictionary Cache Miss Ratio : 0.0524
> >
> > [Shared Pool Usage] Exec Time 0 seconds
> > ~ Total Mb Unused : 30.75
> > ~ Total Mb Used : 273.25
> > ~ Total Mb : 304
> > ~ Shared Pool Percent Used : 89.88
> >
> >
> > [Archive Log Mode Info] Exec Time 0 seconds
> > ! Archiver : STOPPED
> > ! Log Mode : NOARCHIVELOG
> >
> > [Archive Log Info] Exec Time 0 seconds
> > ~ Average Log Switches Per Day : 3.77
> > ~ Hard Drive Storage (in Mb) for this many archive logs : 377.05
> >
> > [Rollback Segments with wait ratios > 1 %] Exec Time 1 seconds
> > ~ None
Received on Tue Nov 05 2002 - 20:40:46 CST
 |
 |