If your intention is to find the amount of memory used
by an Oracle Process at the OS level, use pmap command
in Solaris.
-Ravi.
--- Roger Xu [EMAIL PROTECTED] wrote:
I can use ipcs -am to see the sga at os level. but
I do not see any pga?
Roger Xu
Database Administrator
Dr Pepper
Well, for which process do you want to see PGA? Go to the /proc/$PID
directory and look into the memory maps.
prank
The other way of looking into
PGA would be interpreting process tables from /dev/kmem.
If you know how to do that, you can do something like
dd if=/dev/mem of=`tty`
/prank
On
Please make sure SGA_MAX_SIZE works on your h/w platform. It does not work on AIX as
advertised,
and I think Sun Solaris needs some tweaking (related to ISM use) to make it work.
- Kirti
--- Tanel Poder [EMAIL PROTECTED] wrote:
In 9202, DB_BUBBER_CACHE, SHARED_POOL, LARGE_POOL
and
In 9202, DB_BUBBER_CACHE, SHARED_POOL, LARGE_POOL and JAVA_POOL can be dynamically
altered. But in 901, LARGE_POOL and JAVA POOL are static.
If MAX SGA is less than 128MB then Oracle will use 4MB granule size to
allocate/deallocate memory. For SGA greater than 128M, Oracle granule size is 16MB.
In 9202, DB_BUBBER_CACHE, SHARED_POOL, LARGE_POOL
and JAVA_POOL can be dynamically altered. But in 901, LARGE_POOL
and JAVA POOL are static. If MAX SGA is less than 128MB then Oracle
will use 4MB granule size to allocate/deallocate memory. For SGA greater
than 128M, Oracle granule size is
Raj
First, you may want to consider reading up on virtual memory vs. real
memory. Most modern operating systems use virtual memory. Your program runs
in the virtual memory address space. The O.S. decides which portion of the
virtual memory address space to have in real memory at any given time.
Shouldn't time_waited/100 be time_waited*100?
Regards
Naveen
-Original Message-From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED]Sent: Friday, August 15, 2003 11:19
PMTo: Multiple recipients of list ORACLE-LSubject: Re:
SGAMitchell,
Which portion of the statspack report
by time_waited
/
ttitle off
Jared
Mitchell [EMAIL PROTECTED]
Sent by: [EMAIL PROTECTED]
08/15/2003 09:24 AM
Please respond to ORACLE-L
To:Multiple recipients of list ORACLE-L
[EMAIL PROTECTED]
cc:
Subject:Re
Ya exactly. I had a brain fart.
Got to improve my basic arithmetic :-)
Regards
Naveen
-Original Message-
From: Hately, Mike (LogicaCMG) [mailto:[EMAIL PROTECTED]
Sent: Tuesday, August 19, 2003 3:19 PM
To: Multiple recipients of list ORACLE-L
Subject: RE: SGA
No, time waited
Was a low cache hit ratio the only 'problem'?
Were jobs taking longer than normal?
Were users complaining of a slow system?
Did your average response time shoot up dramatically?
I'm afraid you may have succumbed to the dreaded disease,
CTD, or Compulsive Tuning Disorder.
This is the urge to
Hi Jared
I have reset to previous value and restarted the database anyway. Since I
have 8G Memory and I may set SGA more than 3G. Actually I did at our AIX
SP that set SGA total up to 5G of 8G memory.
Anyway I found a solution on metalink ( Note 115753.1 and 1028623.6. ) to
set SGA to
Please respond to ORACLE-L
To:Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
cc:
Subject:Re: SGA
Hi Jared
I have reset to previous value and restarted the database anyway. Since I
have 8G Memory and I may set SGA more than 3G. Actually I did at our AIX
SP
]
Sent: Friday, August 15, 2003 12:50
PM
Subject: Re: SGA
Mitchell, Which portion of the statspack report suggests
increasing the cache size? Maybe
you could post that portion here. The cause for poor performance needs to be located and
corrected. What are the user
complaints
What's with all this performance stuff? This is a matter of principle. The
man paid for 8 Gig of memory, and BY GOD, the man ought to be about to use
his 8 Gig of memory!
-Original Message-
col event format a35 head 'EVENT NAME'
col total_waits format 999,999,999 head TOTAL|WAITS
Hi DBAs
We have Sun Sparc system (Sun Fire 880) with 8G memeory. We has setting
/etc/system to the max and data buffer catch 1000m and total SGA 1745MB on
our . Oracle 8.1.7.3.0 Server.Since Dat Buffer Cache hit ration
lower then 50% for last a few days, So We decide to increase another
Don't do it.
You should try to avoid paging.
Igor Neyman, OCP DBA
[EMAIL PROTECTED]
-Original Message-
Sai Selvaganesan
Sent: Tuesday, July 15, 2003 11:49 AM
To: Multiple recipients of list ORACLE-L
hi
is it possible to have a sga bigger than the rela
memory available?
suppose i
you will end up doing swap/paging .
-ak
- Original Message -
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Sent: Tuesday, July 15, 2003 9:49 AM
hi
is it possible to have a sga bigger than the rela
memory available?
suppose i have a 1gb ram can i start an
surely i will not do it. but my question is whether it
is possible at all to do it.
will oracle when allocating shared memory space take
virtual memory into consideration or only real memory
into consideration.
thanks
sai
--- AK [EMAIL PROTECTED] wrote:
you will end up doing swap/paging .
If it take only real mem in consideration why would pageing happen at all ?
-ak
- Original Message -
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Sent: Tuesday, July 15, 2003 11:19 AM
surely i will not do it. but my question is whether it
is possible at all to do it.
Hi!
I believe that you can allocate more virtual memory to Oracle than you got
physical. Never cared enough to try, though.
There's a init.ora parameter LOCK_SGA, if you set it to true, then Oracle
tries to hard lock all pages to physical memory, in that case you would
probably get error on
Query V$SGA. Unless you are running Oracle9i, there are no memory leaks in
the SGA, because all allocation of memory for the SGA (prior to Oracle9i)
occurs at instance startup and then at no other time. Just as in crime
investigations, you must have MO: motive and opportunity. No
opportunity,
Hello guru...
Did compilation of package/procedure use SGA segment ?
What kind of parameter setting in ora.ini effect compilation of package ?
Is it possible to increase share-pool to increase the speed of
compilation of package ? Any developer type of configuration setting in for
Yosi, here is the output from v$parameter and
v$system_parameter. As far as I was aware, ismodified will be non-false if an
alter session or alter system command has been carried out. isadjusted will be
non-false if Oracle has dynamically adjusted the parameter.
Show SGA has the following
the diffrence.The correct view to
check size of sga is v$sga.
Hope this will help u.
Thanks
-Seema
From: Hermanto P [EMAIL PROTECTED]
Reply-To: [EMAIL PROTECTED]
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Subject: RE: SGA info.
Date: Wed, 08 Aug 2001 20:51:10 -0800
Raja,
if i do what you do
: Hermanto P [EMAIL PROTECTED]
Reply-To: [EMAIL PROTECTED]
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Subject: RE: SGA info.
Date: Wed, 08 Aug 2001 20:51:10 -0800
Raja,
if i do what you do..the value is same.
i am not sure if you said the value between v$sgastat and v$sga is
different
the diffrence.The correct view to
check size of sga is v$sga.
Hope this will help u.
Thanks
-Seema
From: Hermanto P [EMAIL PROTECTED]
Reply-To: [EMAIL PROTECTED]
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Subject: RE: SGA info.
Date: Wed, 08 Aug 2001 20:51:10 -0800
Raja,
if i do what
If u see all three output you can see the diffrence.The correct view
to check size of sga is v$sga. Hope this will help u.
Thanks
-Seema
From: Hermanto P [EMAIL PROTECTED]
Reply-To: [EMAIL PROTECTED]
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Subject: RE: SGA info.
Date: Wed, 08
4112384 bytes
If u see all three output you can see the diffrence.The correct view
to check size of sga is v$sga. Hope this will help u.
Thanks
-Seema
From: Hermanto P [EMAIL PROTECTED]
Reply-To: [EMAIL PROTECTED]
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Subject: RE: SGA info.
Date
4112384 bytes
If u see all three output you can see the diffrence.The correct view to
check size of sga is v$sga.
Hope this will help u.
Thanks
-Seema
From: Hermanto P [EMAIL PROTECTED]
Reply-To: [EMAIL PROTECTED]
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Subject: RE
Raja,
if i do what you do..the value is same.
i am not sure if you said the value between v$sgastat and v$sga is
different.
SVRMGR select sum(value) from v$sga;
SUM(VALUE)
--
64046072
1 row selected.
SVRMGR select sum(bytes) from v$sgastat;
SUM(BYTES)
--
64046072
1 row
The value fetched by
select sum(bytes) from v$sgastat;
is less than that by other query.
So I am not sure why the values are different. I am having Oracle 8i, 8.1.5 on HP11.
rgds,
raja
--
On Wed, 08 Aug 2001 20:51:10
Hermanto P wrote:
Raja,
if i do what you do..the value is same.
i am
It's a bug. The bug fix is out there. It seems to come from
connect/disconnect.
-Original Message-
Sent: Tuesday, July 03, 2001 9:57 AM
To: Multiple recipients of list ORACLE-L
Hello, dear listers,
I've got 8.1.7.1 installation on RH6.2 here.
Runs smoothly, but I noted strange
Thank you, Kimberly,
I've doublechecked the bug (did you mean this one 8171 [BUG:1240484]
Shared server memory leak on repeated connect/disconnect ?), it doesn't
seem applicable to my case - I dont use MTS, dedicated only.
Thank you anyway, I feel more confidence to iTAR now, this is very
I found this in the metalink today since we plan to upgrade to 8.1.7.1:
Try this thread:
http://metalink.oracle.com/metalink/plsql/ml2_documents.showFOR?p_id=115437.
996p_showHeader=1p_showHelp=1
-Original Message-
Sent: Tuesday, July 03, 2001 3:11 PM
To: Multiple recipients of list
Actually its bug 1397604
-Original Message-
Sent: Tuesday, July 03, 2001 12:11 PM
To: Multiple recipients of list ORACLE-L
Thank you, Kimberly,
I've doublechecked the bug (did you mean this one 8171 [BUG:1240484]
Shared server memory leak on repeated connect/disconnect ?), it doesn't
I was told by Metalink that I do not have access to this section on
Metalink. How did you searched for it?
Alex Hillman
-Original Message-
Sent: Tuesday, July 03, 2001 3:35 PM
To: Multiple recipients of list ORACLE-L
I found this in the metalink today since we plan to upgrade to
Alex,
Full link is
http://metalink.oracle.com/metalink/plsql/ml2_documents.showFOR?p_id=115437.
996p_showHeader=1p_showHelp=1
It was wrapped over in the original posting.
Also, I found many references to this bug searching ML for 1397604.
Gonna set _db_handles_cached = 0 for now. And
Hi!
Vadim Gorbounov writes:
Thank you, Kimberly,
I've doublechecked the bug (did you mean this one 8171 [BUG:1240484]
Shared server memory leak on repeated connect/disconnect ?), it doesn't
seem applicable to my case - I dont use MTS, dedicated only.
Thank you anyway, I feel more
200MB? I realize simply adding memory is not the solution for fixing a
poorly tuned database, but on the other hand, if you have a lot of memory
why not use it?
- Greg
--
Hi Jared,
I think like Greg . Why not use 2-4 gb of 8 gb memory for SGA. ?
If It has bad results
I had read from a paper that . NEVER EXCEED 55% of totaL memORY FOR NT . but
I KNOW THAT YOU CAN NOT EXCEED 2 GB TOO.
- Original Message -
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Sent: Monday, June 25, 2001 10:53 PM
Does anyone have an idea of what percentage of
My database will be growing 20 GB per
year.
Anyway . Forget it .
Thank you All.
Bunyamin
- Original Message -
From:
Christopher
Spence
To: Multiple recipients of list ORACLE-L
Sent: Monday, June 25, 2001 6:32 PM
Subject: RE: SGA QUESTION
Who
knows
Kevin and John,
The original limit was 2GB on NT, but since service
pack 3 and above this can be extended to 3GB by
changing a few settings. There is also a special
driver that can be obtained to allow access up to 8GB.
I have never tried this driver since I don't have any
boxes with 4GB, but
Hi Jared,
I think like Greg .
Oh, we're in trouble now
- Greg
--
Please see the official ORACLE-L FAQ: http://www.orafaq.com
--
Author: Greg Moore
INET: [EMAIL PROTECTED]
Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051
San Diego, California-- Public
HERE IS THE KNOWLEDGE,
my database is nearly 700 tables where 2 of it is 3 gb growing per year.
My total disk is 400 GB.
Total number physical users which will connect to database is 1000 but will
be
2 for 2 years.
Total number of database users is 30.
Oracle is on NT.
Oracle Version is
I believe that you can have 8 Gig's, but don't you have to apply a patch in
order for it to work?
KK
-Original Message-
Dayal
Sent: Monday, June 25, 2001 6:36 AM
To: Multiple recipients of list ORACLE-L
Oracle is on NT.
My RAM is 8 GB.
Is it a good (VALID) combination ;-)
Rajesh
Who
knows.
Perhaps your database is 100K perhaps it is 1000Pb.
Perhaps it is DSS, DW, or even OLTP.
I
would say use a 8Gb sga and buy more memory, should be perfect for what your
trying to do.
"Walking on water and developing software from
a specification are easy if both are frozen."
Hi all,
If he has 8GB of memory, why not allocate 4GB to the data
buffer instead of
I don't remember the limit, but there *is* an upper limit on the amount of
memory that a single process can address in NT (was it 2Gb?). Since the
architecture of Oracle on NT is a single-process-multi-threaded
Hello,
The upper limit for a server running NT is 2gb, because usually only 4gb of
memory can be installed. However, I believe that with the patch installed,
enabling NT to go above 4gb of ram, it will also increase the upper limit.
KK
-Original Message-
Kanagaraj
Sent: Monday,
Does anyone have an idea of what percentage of total memory should be
reserved for NT? Say I had 600m of memory available. How big could my SGA
be and still have NT run properly?
Ron Smith
Database Administration
[EMAIL PROTECTED]
-Original Message-
Sent: Monday, June 25, 2001 2:33
You suck up half of that memory easy for Oracle and NT would run just fine,
as long as there aren't a whole bunch of other applications running on the
server!
Kev
-Original Message-
L.
Sent: Monday, June 25, 2001 3:54 PM
To: Multiple recipients of list ORACLE-L
Does anyone have an idea
Message-
From: Jared Still [SMTP:[EMAIL PROTECTED]]
Sent: Saturday, June 23, 2001 3:30 PM
To: Multiple recipients of list ORACLE-L
Subject: Re: SGA QUESTION
Hoo boy! That's a pretty healthy SGA you got there.
Why do you think you need to dedicate that much RAM
to your SGA
Hoo boy! That's a pretty healthy SGA you got there.
Why do you think you need to dedicate that much RAM
to your SGA?
Why not start with something more reasonable, like say,
200 meg for shared pool and 200 more for the database
buffers?
A large shared pool can actually impede performance.
Jared,
If he has 8GB of memory, why not allocate 4GB to the data buffer instead of
200MB? I realize simply adding memory is not the solution for fixing a
poorly tuned database, but on the other hand, if you have a lot of memory
why not use it?
- Greg
--
Please see the official ORACLE-L FAQ:
On Saturday 23 June 2001 15:15, Greg Moore wrote:
If he has 8GB of memory, why not allocate 4GB to the data buffer instead of
200MB? I realize simply adding memory is not the solution for fixing a
poorly tuned database, but on the other hand, if you have a lot of memory
why not use it?
This Metalink doc has what you're looking for
http://metalink.oracle.com/metalink/plsql/ml2_documents.showDocument?p_database_id=NOTp_id=1008866.6
Harsh Agrawal wrote:
Hi,
Can u help me to knoe what are the initDB.ora parameters contributing
these ?
Pl. give one to one or one to
56 matches
Mail list logo