Notes in-line
Regards
Jonathan Lewis
http://www.jlcomp.demon.co.uk
The educated person is not the person
who can answer the questions, but the
person who can question the answers -- T. Schick Jr
One-day tutorials:
http://www.jlcomp.demon.co.uk/tutorial.html
Three-day seminar:
see
Hi Yong
I agree that It's not the ratio that needs condemning, it's
the advice about... What I disagree is the wrong educational
tool people on public forums have recently used again and
again to show the inadequacy of the BCHR tuning method.
The thing is most people who have used this
#1. what is the difference in the 4 BCHR's in 9i?
#2. How do you determine whether your buffer cache is sized properly?
- Original Message -
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Sent: Saturday, December 27, 2003 9:34 AM
Hi Yong
I agree that It's not the ratio
Hello list,
Can someone please explain to me why the following order by clauses are
valid and yield the same results :
select empno, deptno from emp
order by sqrt (1) ;
and
select empno, deptno from emp
order by sqrt ( 3.14234 ) ;
The docs say that in the order by clause you could specify only
The credit for that goes to Gary Dodge. It has been his email signature for
better than 10 years...
Building tomorrow's legacy systems today - one crisis at a time
Gene Fosnight (a successfully retired Oracle consultant) had an email
signature that was, if anything, even better:
Look, listen,
Also, consider that any single serial session will never get more than 5%
of
pga_aggregate_target. For parallel operations, total is limited to 30%.
The maximum can be controlled using _smm_max_size parameter which states
how many kilobytes a serial session can use for its workarea operations.
Because both expressions evaluate to
1and cause the output to be sorted by the first column? Oracle seems
to ignore the values after the decimal point.
SQL select a, c from x
2 where rownum 10 3 order by 1.7;
A C--
- 4861 Y
4862 N 4863
Y 4864 N
4865 Y 4866
N 4867 Y
4868 N
Hi!
You may want to set _smm_max_size parameter to higher value (see my recent
post about pga_aggregate_target), but I think setting your sessions pga
management temporarily to manual + setting sort_area_size, as Jared and
Jonathan already recommended, is better solution for your current issue.
#1. what is the difference in the 4 BCHR's in 9i?
What do you mean by that? Are you refferring to different buffer pool stats
here?
#2. How do you determine whether your buffer cache is sized properly?
Your application meets it's defined operational response time and throughput
constraints.
I think the JDBC driver has connection pooling also, but not session
pooling. OCCI might have both. But I haven't checked.
OCCI is AFAIK just an OO wrapper around OCI so it should have every
functionality that OCI has.
Tanel.
--
Please see the official ORACLE-L FAQ: http://www.orafaq.net
Saturday, December 27, 2003, 11:14:25 AM, [EMAIL PROTECTED] ([EMAIL PROTECTED]) wrote:
bnini Hello list, Can someone please explain to me why the
bnini following order by clauses are valid and yield the
bnini same results :
bnini select empno, deptno from emp
bnini order by sqrt (1) ;
The
Thanks Tanel. I wasn't aware of those parameters.
-Original Message-
From: Tanel Poder [mailto:[EMAIL PROTECTED]
Sent: Sat 12/27/2003 11:59 AM
To: Multiple recipients of list ORACLE-L
Cc:
Subject:Re: pga_aggregate_target
Also, consider that any single serial
There are two things that I can theorize could cause this problem.
1) Oracle bug (have you logged a tar with OWS?)
2) The block sampled is below the high water mark, but does not contain any data. I
don't know the
exact sampling algorithm, but this seems to be the 2nd simplest explanation (bug
JP, Juan, and Prem,
Prem is correct. Standby is certainly a feature of SE, just not the
managed recovery and SQL*Net log shipping parts, which only come with
EE. Essentially, Standby Database features in SE is just like Standby
Database features from 7.3 through 8.0.
I have some shell scripts
Note that when finding sessionid, querying from
v$mystat where rownum=1 is faster than selecting sys_context from dual, because
dual access requires full table scan on it, which means 3-5 LIOs, depending on
version, buffer pins, latches etc..
Even when you don't select sys_context from
I agree that this difference might be only because
sqlnet is much more "fat" that ICMP.
But anyway, could some overhead be added be because
the failover load balancing clauses that require extra work?
Also, if listener logs every connection, this might
add some extra IO time as well (if
Title: Exporting a partition with transport tablespace
Make sure that you use novalidate and exchange
indexes as well - that way Oracle won't make unnecessary work.
Also, you should export the metadata at the time
when required partitions are exchanged to "transport tables".
Tanel.
-
#2. How do you determine whether your buffer cache is sized
properly?
Your application meets it's defined operational response time
and throughput constraints.
Actually that is an interesting answer, since it doesn't answer the question
asked 'How do I know if my buffer cache is sized
Good news !
That bug has been fixed in 9.2.0.4
Regards
Jonathan Lewis
http://www.jlcomp.demon.co.uk
The educated person is not the person
who can answer the questions, but the
person who can question the answers -- T. Schick Jr
One-day tutorials:
Performance 'problems' are dependent on what the marketing department gets
in the SLA.
So if your marketing guys negotiate very strict response time requirements
and you dont meet them, then you have a performance problem. How do I know
if my buffer cache is having any effect on that?
-
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of Ryan
Sent: 27 December 2003 21:59
To: Multiple recipients of list ORACLE-L
Subject: Re: Correct way to accuse BCHR tuning method (Was: Hit ratio)
Performance 'problems' are dependent on what the
#2. How do you determine whether your buffer cache is sized
properly?
Your application meets it's defined operational response time
and throughput constraints.
Actually that is an interesting answer, since it doesn't answer the
question
asked 'How do I know if my buffer cache is
Jonathan,
Which exact behaviour were you talking about?
Tanel.
- Original Message -
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Sent: Saturday, December 27, 2003 11:34 PM
Good news !
That bug has been fixed in 9.2.0.4
Regards
Jonathan Lewis
Hi Jonathan,
I'm not sure what you really think about this new feature!
Are you saying that Oracle is capable now of releasing the extra memory
something it was not capable of before?
If yes, then what does it have to do with the work policy?
I see this feature useful (not really) for a
It would appear that the values for _smm_max_size and _smm_px_max_size
are specified in K, though not explicitly.
In a test 9.2.0.4 database:
pga_aggregate_target = 25165824
_smm_max_size = 1228
_smm_px_max_size = 7371
Anyone know this for sure?
Jared
On Sat, 2003-12-27 at 08:59, Tanel
Quite a surprise to find this on the front page of OTN
http://otn.oracle.com/pub/articles/dulaney_sed.html
Jared
--
Please see the official ORACLE-L FAQ: http://www.orafaq.net
--
Author: Jared Still
INET: [EMAIL PROTECTED]
Fat City Network Services-- 858-538-5051
Hi,
A simple rollback segment question. If I do not set a optimal thencan I assume following
1)rollback segment will not shrink.
2)It will keep growing till tablespace fills up.
3)And there will be no 'snapshot too old' errors?
ThanksA Joshi [EMAIL PROTECTED] wrote:
Hi,
One rollback
27 matches
Mail list logo
