SGA size is 1 GB free but oracle starts physical read/write,why ?

Dear Friends ,
In my production server I observe that it takes the the data from the HDD where my SGA is still 1 GB free .
As far I know , In oracle , when the allocated SGA size is filled up then it reads the data from the HDD which is called "physical read/write " . I am using oracle 10g database . In performance tab(From Oracle 10g OEM) , when I see that oracle starts to read the physical data , at that moment I run the following query :
select sum(bytes)/1024/1024 " SGA size used in MB"
from v$sgastat
where name!='free memory';
and I found that still 1 GB memory are still free in the SGA .
In this moment , I want to know , is it possible to start the physical read / write while the SGA size are still free ?
Edited by: shipon_97 on Mar 24, 2009 10:45 AM

shipon_97,
Your understanding is wrong. From your previous posts it appears you never read the concepts manual.
Begging on my knees PLEASE READ IT and stop asking doc questions.
Sybrand Bakker
Senior Oracle DBA

Similar Messages

Oracle coherence first read/write operation take more time

I'm currently testing with oracle coherence Java and C++ version and from both versions for writing to any local or distributed or near cache first read/write operation take more time compared to next consecutive read/write operation. Is this because of boost operations happening inside actual HashMap or serialization or memory mapped implementation. What are the techniques which we can use to improve the performance with this first read/write operation?
Currently I'm doing a single read/write operation after fetching the NamedCache Instance. Please let me know whether there's any other boosting coherence cache techniques available.

In which case, why bother using Coherence? You're not really gaining anything, are you?
What I'm trying to explain is that you're probably not going to get that "micro-second" level performance on a fully configured Coherence cluster, running across multiple machines, going via proxies for c++ clients. Coherence is designed to be a scalable, fault-tolerant, distributed caching/processing system. It's not really designed for real-time, guaranteed, nano-second/micro-second level processing. There are much better product stacks out there for that type of processing if that is your ultimate goal, IMHO.
As you say, just writing to a small, local Map (or array, List, Set, etc.) in a local JVM is always going to be very fast - literally as fast as the processor running in the machine. But that's not really the focus of a product like Coherence. It isn't trying to "out gun" what you can achieve on one machine doing simple processing; Coherence is designed for scalability rather than outright performance. Of course, the use of local caches (including Coherence's near caching or replicated caching), can get you back some of the performance you've "lost" in a distributed system, but it's all relative.
If you wander over to a few of the CUG presentations and attend a few CUG meetings, one of the first things the support guys will tell you is "benchmark on a proper cluster" and not "on a localised development machine". Why? Because the difference in scalability and performance will be huge. I'm not really trying to deter you from Coherence, but I don't think it's going to meet you requirements when fully configured in a cluster of "1 Micro seconds for 100000 data collection" on a continuous basis.
Just my two cents.
Cheers,
Steve
NB. I don't work for Oracle, so maybe they have a different opinion. :)

I purchased a new iPad mini, downloaded iWork and iLife for free but Pages still shows $9.99, why? How can I download Pages for free too?

That's basically my question, please help! u_u

Pages is one of the apps included in iWork.
How to get all the iWork apps, iPhoto, and iMovie for free on an eligible iPhone or iPad
http://www.imore.com/how-get-all-iwork-apps-iphoto-and-imovie-free-eligible-ipho ne-or-ipad
About Free Apple Apps for iOS 7 compatible devices
http://support.apple.com/kb/HT5913
 Cheers, Tom

Passport will mount on my desktop, but cannot access drive( read/write/copy/move) Yosemite MBP 2007

I have an older MacBook Pro (2007-2008) running the most current version of Yosemite. I just purchased my drive recently, but didn't try to use it till the other day. The drive mounts on my desktop, but I cannot copy files to it at all. When I try to do this, I get no error messages of any type. It just doesn't complete the action when I try to do so. I've tried "dragging" files to the drive. I've tried copy and paste. Nothing seems to work. I bought this drive for my Mac laptop exclusively, for file storage and transfer, etc. I do not intend to use it as a Time Machine backup drive. Basically, I want to be able to move files from any computer (Mac or PC), to this drive, for use, editing, etc. on my Mac. This drive is currently empty. No files on it. This was my 1st time use attempt. Is this possibly just a formatting issue? I've not changed anything on the drive. I literally took it out of the packaging, and plugged it into my Mac. It has not been used (yet) on a PC. Suggestions? Thank you, Dave

I can format my drive, but I want to retain the cross platform in doing so. I just can't remember how to do it. I do not know what the drive is formatted to, out of the box. Is that why I cannot write any files from my mac, to that drive? Again, it will mount on my Mac desktop, but wont let me do anything else. If I format, I just want to make sure I can copy files from my PC, if necessary, and still have access to those files on my Mac. And, visa versa... Thanks again, Dave

PDF opens and works correctly with Windows 8 Reader, but not with Adobe Reader. Why?

Hi,
My sister-in-law has this PDF order form from one of her company's distributors that opens and allows editing of specific fields.
The only application that opens the file correctly (visually and functionally) is the Windows 8 Reader. If I open the file in Adobe Reader or with the Mac viewer things do not line up, fonts are too large, and when editing fields sometimes symbols show up and sometimes way over-sized numbers show up.
This behavior has been observed on both a Windows 8.1 PC as well as on a Mac Book Air.
Can someone please tell me why this is happening? Is there something that needs to be configured differently?
Thanks,
Matt

This is what I see when I open that PDF in Adobe Reader 11.0.07:
Can you tell me which parts are not correct? Obviously the Submit button is a bit strange.

Uniform SGA sizes or not?

Our shop is currently running Oracle Applications 11.5.10.2 on a single node database(10.2.0.2) on Sun Solaris10. We are planning to migrate to RAC(10.2.0.3) and have started work on our development servers. We are planning our production architecture and have a question related to SGA sizing between nodes.
We our trying to get a better understanding of what our SGA sizes on each of our nodes should be if we have one large capacity node and several smaller capacity nodes. By capacity I mean the size of memory and number of CPUs. The biggest concern is the size of memory. On our large box we will have 48G and on each subsequent node we will have 32G. The SGA on the big node is currently about 12G and the PGA 8G. Combined with other memory requirements we can consume 28-30G on our server. My questions are: What is Oracle's recommendation for sizing the SGAs on the smaller nodes? If we size them based on our large node would we run into memory shortages on the smaller nodes? Does RAC require a uniform SGA size across all nodes? Thanks.

Oracle does not require the same SGA on all nodes. All machines that are intended to share the same workload will typically have similar SGA sizing. However, machines that are intended to carry different workloads will typically have different SGA configurations.
The SGA is node specific and is defined by the workload expected on that node. Realize that the SGA has a number of pools, the major ones being 'data buffer' and 'shared pool'.
In very crude terms, the shared pool is used to handle the SQL & PL/SQL, and the buffer is used to handle the data being affected by the SQL and PL/SQL.
If a node is designated to handle specific workload ("this is the batch machine", "this is the backup machine", "these machines are for the customer service team's web screens") the SQL thrown at the node will be potentially different fromn any other node and it should be sized for that workload. This is standard 'performance tuning'.
Workload segregation is typically accomplished using the 'Service' capability of the database and listener.
IOW, RAC does not eliminate the DBA's need to tune based on workload. The tuning is done at the instance level.

Apps that say free but with a subscription price added?

When you have an App such as a magazine, or news like The New York Times with a monthly subscription. When you download the app for free are you getting charged for that monthly fee without knowing it? I ask because the app says free download but is not so clear on if you subscribed on a monthly subscription. So is the buyer being fooled or is it really free?

That's referred to as an "In-App" purchase...
Help here ...
iTunes Store: About In-App Purchases
iTunes Store: Purchasing and managing auto-renewing subscriptions
The app to read the content is free but not the subscription.

ORACLE SGA size

We bought a server for oracle data warehousing which has got 200GB RAM. Can I set set my SGA_TARGET to 100GB. Is it true the more SGA size result in better performance.
Thaks
jeevan

Hi Jeevan,
I would not set sga_target, it can cause RAM thrashing . . .
In an article titled Automatic Shared Memory Management we see:
http://www.oracle.com/technology/pub/articles/10gdba/week17_10gdba.html
Some pools in SGA are not subject to dynamic resizing, and must be specified explicitly. Notable among them are the buffer pools for nonstandard block sizes and the non-default ones for KEEP or RECYCLE. If your database has a block size of 8K, and you want to configure 2K, 4K, 16K, and 32K block-size pools, you must set them manually.
Their sizes will remain constant; they will not shrink or expand based on load. You should consider this factor when using multiple-size buffer, KEEP, and RECYCLE pools.
In addition, log buffer is not subject to the memory adjustment—the value set in the parameter log_buffer is constant, regardless of the workload. ( In 10g, a new type of pool can also be defined in the SGA: Streams pool, set with parameter streams_pool_size. This pool is also not subject to automatic memory tuning.)
Is it true the more SGA size result in better performance. Yes, but it may not be much, or it could be alot! It depends.
On a dedicated server, your goal is to allocate all the RAM, without causing swapping. I have my notes here:
http://www.dba-oracle.com/art_dbazine_ram.htm
Hope this helps . . .
Donald K. Burleson
Oracle Press author
Author of "Oracle Tuning: The Definitive Reference"
http://www.rampant-books.com/book_2005_1_awr_proactive_tuning.htm

How to decrease SGA size in oracle 9i

hi,
how to decrease SGA size .. now SGA size=400m i want to drecrease to 150m
Thanks,
Mohammed

Why do you want to do this? Is your system low on memory?
Perhaps your DB is on a laptop with
only a small amount of memory?
Well an enterprise database server set of
thread inside a complex service would need
a high powered laptop with large amounts
of memory.
Or perhaps your UAT server has 10 databases
on it all competing for resource?The culprit may not be Oracle.
For example (as you have not told us your operating system, windows linux unix?) IBM's AIX unix operating system comes configured optimally for managing file caching in the OS memory; as Oracle also caches the blocks read from disk sometimes the file system caching is redundant and ends up using memory better allocated to Oracle. In that case you tune the Unix kernel to reduce memory allocated to the file system caching. The default settings allow IBM AIX to take up to 80% of system memory for caching bad if you have allocated 40% to the SGA already.........
The really cool thing about AIX is you can do this on the fly and you don't need to reboot so when you go live you juggle until you get an optimal setting and then monitor.
In the end you may just have to add memory.

How can i increase sga size in oracle 10g

Hello friends
how can i increase my sga size in oracle 10g
Regards
Vicky
Edited by: Vignesh Chinnasamy on 31-Jul-2012 02:28

HI
**SQL> Show parameter sga ;**
NAME                                 TYPE        VALUE
lock_sga                             boolean     FALSE
pre_page_sga                         boolean     FALSE
sga_max_size                         big integer 2G
sga_target                           big integer 2G
**SQL> show parameter memory;**
NAME                                 TYPE        VALUE
hi_shared_memory_address             integer     0
shared_memory_address                integer     0
**[root@mte ~]# ulimit -a**
core file size          (blocks, -c) 0
data seg size           (kbytes, -d) unlimited
file size               (blocks, -f) unlimited
pending signals                 (-i) 1024
max locked memory       (kbytes, -l) 32
max memory size         (kbytes, -m) unlimited
open files                      (-n) 1024
pipe size            (512 bytes, -p) 8
POSIX message queues     (bytes, -q) 819200
stack size              (kbytes, -s) 10240
cpu time               (seconds, -t) unlimited
max user processes              (-u) 278528
virtual memory          (kbytes, -v) unlimited
file locks                      (-x) unlimited
*[root@mte ~]#*

How to increase SGA size in Oracle 9i release 9.0.1.1.1

Hi,
We have Oracle 2000 server on a intel machine (2.8GHz processor and 2 GB RAM). Virtual memory is :- 3GB.
We have a Oracle 9.0.1.1.1 running on this machine. What are the parameters that I need to set to increase the SGA size. We want a large SGA size with enough space for shared pool.
Thanks in advance.
--Malay

Malay,
First Check out the RAM utilization for the OS Kernel and remaining to the applications and then to SGA.
We have Oracle 2000 server on a intel machine (2.8GHz processor and 2 GB RAM). Virtual memory is :- 3GB.I suppose its Windows 2000 Server!!!
& What's the size of the Database?
all these things allow you to decide how "large" you can increase the SGA

Reduced SGA_TARGET, but SGA size not changing?

I reduced the sga_taget from 1536M to 512M:
alter system set sga_target = 500M scope = memory;
System altered.
select VERSION from v$instance;
VERSION
10.2.0.3.0
show parameter sga
NAME TYPE VALUE
lock_sga boolean FALSE
pre_page_sga boolean FALSE
sga_max_size big integer 1536M
sga_target big integer 512M
But the real memory still showing the original value
show sga
Total System Global Area 1610612736 bytes
Fixed Size 2030456 bytes
Variable Size 1509950600 bytes
Database Buffers 83886080 bytes
Redo Buffers 14745600 bytes
why is that while it's a dynamic parameter?
Thanks a lot for any help.
Edited by: user10484253 on May 13, 2011 8:36 AM
Edited by: user10484253 on May 13, 2011 8:39 AM
Edited by: user10484253 on May 13, 2011 8:41 AM

I would suggest you to check v$sgastat to find out the exact SGA memory you are using currently instead of using "SHOW SGA" when when you set SGA_MAX_SIZE & SGA_TARGET initialization parameters.
Below is a sample output from one of my test dbs. As you can see below my SGA size is only 1GB.
SHOW SGA shows 2GB thats because I have set SGA_MAX_SIZE to 2gb ( which only means that I can grow my sga up till 2 gig , it may not be my current sga size).
you can try increasing or decreasing SGA_TARGET and check memory usage on OS level to see the difference.
SQL>show parameter sga
NAME                                 TYPE        VALUE
lock_sga                             boolean     FALSE
pre_page_sga                         boolean     FALSE
sga_max_size                         big integer 2000M
sga_target                           big integer 1008M
SQL>show sga
Total System Global Area 2087780352 bytes
Fixed Size                  2155336 bytes
Variable Size            1744833720 bytes
Database Buffers          318767104 bytes
Redo Buffers               22024192 bytes
SQL>select name, round(sum(mb),1) mb
2        from (
3      select case when name = 'buffer_cache' then 'db_cache_size'
4                  when name = 'log_buffer'   then 'log_buffer'
5                  else pool
6              end name,
7              bytes/1024/1024 mb
8                   from v$sgastat
9           )group by name
10 /
NAME                  MB
db_cache_size        304
java pool            128
large pool            16
log_buffer            21
shared pool          528
                     2.1
6 rows selected.
SQL> -- V$SGA_DYNAMIC_FREE_MEMORY: Information about the amount of SGA memory available for future dynamic SGA resize operations.
SQL>select * from V$SGA_DYNAMIC_FREE_MEMORY;
CURRENT_SIZE
1040187392- Krishna

SGA Size for 8.1.7.4 32 bit? , some Interview Questions

Hi buddies,
I got some interview questions, might be simple for geeks in DBA. I am in need of answers. Could anyone help me.
Thanks,
Raaj
1) Does windows NT support direct I/O?
Answer: Choose one of the answers that apply
A: No, only AIO
B: Yes, depending on hardware.
C: Yes.
D: No.
2) Can you take a coldbackup from solaris and use it on windows NT?
Answer: Choose one of the answers that apply
A: Yes.
B: Yes if RMAN backup performed from NT server.
C: Yes, after running RMAN convert.
D: No.
3) All of the following will alter the number of checkpoints that occur in one hour on the database, except one. Which is it?
Answer: Choose one of the answers that apply
A: Decreasing tablespace size
B: Decreasing size of redo log members
C: Setting LOG_CHECKPOINT_INTERVAL greater than the size of the redo log file
D: Setting LOG_CHECKPOINT_TIMEOUT to zero
4) The DBA is attempting to back up the Oracle database control file. After
issuing the ALTER DATABASE BACKUP CONTROLFILE TO TRACE command, where can the DBA find the backup control file creation materials Oracle created for him or her ?
Answer: Choose one of the answers that apply
A: USER_DUMP_DEST
B: LOG_ARCHIVE_DEST
C: CORE_DUMP_DEST
D: BACKGROUND_DUMP_DEST
5) What is the most important action a DBA must perform after changing the database from NOARCHIVELOG TO ARCHIVELOG?
Answer: Choose one of the answers that apply
A: Shutdown normal and restart the database
B: Perform a full logical database backup
C: Perform a full offline database backup
D: Manually switch the log files
6) Which of the following choices lists an ALTER USER option that can be executed by the user herself or himself?
Answer: Choose one of the answers that apply
A: DEFAULT TABLESPACE
B: IDENTIFIED BY
C: TEMPORARY TABLESPACE
D: PROFILE
7) You need to view the initialization parameter settings for your Oracle
database. Which of the following choices does not identify a method
you can use to obtain values set for your initialization parameters?
Answer: Choose one of the answers that apply
A: Issue SELECT * FROM DBA_PARAMETERS; from SQL*Plus
B: Issue SELECT * FROM V$PARAMETER; from SQL*Plus
C: Issue SHOW PARAMETERS from Server Manager
D: Use OEM Instance Manager
8) As a result of a media failure, the current online redo log group is corrupted, the database crashes, as the current online group is inaccessible. Which type of incomplete recovery are you most likely to perform ?
Answer: Choose one of the answers that apply
A: Change-based
B: Time-based
C: Recovery using a backup control file
D: Cancel-based
9) User SNOW executes the following statement: SELECT * FROM EMP. This
statement executes successfully, and SNOW can see the output. Table
EMP is owned by user REED. What object would be required in order for
this scenario to happen ?
Answer: Choose one of the answers that apply
A: User SNOW would need the role to view table EMP.
B: User SNOW would need the privileges to view table EMP.
C: User SNOW would need a synonym for table EMP.
D: User SNOW would need the password for table EMP.
10) Which one of the following statements is true?
Answer: Choose one of the answers that apply
A: The request queue is common, and the response queue is different for all the dispatchers.
B: The request queue and response queue are different for all the dispatchers.
C: The request queue is different, and response queue is common for all the dispatchers.
D: The request queue and response queue are common for all the dispatchers.
11) What is the largest SGA size for 8.1.7.4 32 bit?
Answer: Choose one of the answers that apply
A: approximately 2GB
B: approximately 3.5GB
C: approximately 4GB
D: approximately 8GB
E: approximately 16GB
12) The DBA is about to perform some administrative tasks. Specifying the
OPTIMAL parameter has which of the following appropriate uses?
Answer: Choose one of the answers that apply
A: Limiting concurrent users
B: Limiting concurrent transactions
C: Limiting growth of rollback segments
D: Limiting growth of tables
13) If the DBA wants to find information about how often transactions are
wrapping transaction information between multiple rollback segment
extents, where would the DBA look to find that information?
Answer: Choose one of the answers that apply
A: DBA_ROLLBACK_SEGS
B: V$ROLLSTAT
C: V$ROLLNAME
D: DBA_SEGMENTS
14) You have 30 rollback segments in your database, for which
TRANSACTIONS_PER_ ROLLBACK_SEGMENT is set to 49 and
TRANSACTIONS is set to 1000. During periods of heavy usage, about how many rollback segments will be actively used by Oracle?
Answer: Choose one of the answers that apply
A: 50
B: 60
C: 20
D: 30
15) The DBA has a table created with the following statement:
CREATE TABLE EMPL
(EMPID NUMBER(10),
LASTNAME VARCHAR2(40),
RESUME LONG RAW);
The DBA attempts to issue the following statement:
ALTER TABLE EMPL
ADD ( PERF_APPRAISE LONG);
What happens?
Answer: Choose one of the answers that apply
A: The statement succeeds.
B: The statement succeeds, but column is added as VARCHAR2.
C: The statement fails.
D: The statement adds a disabled constraint.
16) The primary key of the EMP table has three columns, EMPID, LASTNAME,
and FIRSTNAME. You issue the following SELECT statement:
SELECT * FROM EMP WHERE LASTNAME = 'HARRIS' AND FIRSTNAME = 'BILLI'
AND EMPID = '5069493';
Where would you look to see if this query will use the index associated
with the primary key?
Answer: Choose one of the answers that apply
A: DBA_IND_COLUMNS
B: DBA_TAB_COLUMNS
C: DBA_INDEXES
D: DBA_CLU_COLUMNS
17) You are configuring your index to be stored in a tablespace. Which of the
following storage parameters are not appropriate for indexes?
Answer: Choose one of the answers that apply
A: OPTIMAL
B: INITIAL
C: PCTINCREASE
D: NEXT
18) You need to set up auditing in an order entry and product shipment
application so that when the ORDER_STATUS column in the ORDERS
table changes to SHIPPED, a record is placed in a special table associated
with a part of the application that gives sales representatives a daily list
of customers to call on a follow-up to make sure the customer is satisfied
with the order. Which of the following choices represents the best way
to perform this auditing?
Answer: Choose one of the answers that apply
A: Statement auditing
B: Object auditing
C: Audit by access
D: Value-based auditing
19) Information in the buffer cache is saved back to disk in each of the
following situations except one. In which situation does this not occur?
Answer: Choose one of the answers that apply
A: When a time-out occurs
B: When a log switch occurs
C: When the shared pool is flushed
D: When a checkpoint occurs
20) In order to allow remote administration of users and tablespaces on an Oracle database, which of the following types of files must exist in the database?
Answer: Choose one of the answers that apply
A: Password file
B: Initialization file
C: Datafile
D: Control file
E: Nothing, SYSDBA privileges are not required for these actions.
21) You are planning the storage requirements for your database. Which of the following is an effect of maintaining a high PCTFREE for a table?
Answer: Choose one of the answers that apply
A: Oracle will manage filling data blocks with new records more actively.
B: Oracle will manage filling data blocks with new records less actively.
C: Oracle will leave more space free in data blocks for existing records.
D: Oracle will leave less space free in data blocks for existing records.
22) You manage database access privileges with roles where possible.
You have granted the SELECT_MY_TABLE role to another role, called
EMP_DEVELOPER. To view information about other roles that may be
granted to EMP_DEVELOPER, which of the following dictionary views
are appropriate?
Answer: Choose one of the answers that apply
A: DBA_ROLE_PRIVS
B: DBA_TAB_PRIVS
C: USER_SYS_PRIVS
D: ROLE_ROLE_PRIVS
23) In order to set your SQL*Plus session so that your NLS_DATE_FORMAT
information is altered in a specific way every time you log into Oracle,
what method would be used?
Answer: Choose one of the answers that apply
A: Setting preferences in the appropriate menu option
B: Creating an appropriate LOGIN.SQL file
C: Issuing the ALTER USER statement
D: Issuing the ALTER TABLE statement
24) You create a sequence with the following statement:
CREATE SEQUENCE MY_SEQ
START WITH 394
INCREMENT BY 12
NOMINVALUE
NOMAXVALUE
NOCACHE
NOCYCLE;
Two users have already issued SQL statements to obtain NEXTVAL, and
four more have issued SQL statements to obtain CURRVAL. If you issue a
SQL statement to obtain the NEXTVAL, what will Oracle return?
Answer: Choose one of the answers that apply
A: 406
B: 418
C: 430
D: 442

1.-
2.c
3.a
4.a
5.c
6.b
7.a
8.d
9.b
10.a -
11.a
12.c
13.b
14.d
15.c
16.a -
17.a
18.d
19.c
20.a
21. -
22.d
23.b
24.?
hope it helps u.
Thanks
Kuljeet

SGA size post cloning

Hi all,
I have cloned 11.5.10.2 ebs however after cloning i discovered sga size doesn't match the prod sga size.
How could i change sga size exact to prod sga.
I am using rhel 5.2 with ebs 11.5.10.2
Thanks

I typed free -m & to my surprise
                                 total       used       free     shared    buffers     cached
Mem:                        32497      28784       3712          0        309      28098
-/+ buffers/cache:        376      32120
Swap:                        8189          0       8189
this is a new linux server currently db is down & apps services as well down. still 28gb is being used.
top command output is below
top - 12:41:25 up 8 days, 19:46, 4 users, load average: 0.00, 0.00, 0.00
Tasks: 113 total,   1 running, 112 sleeping,   0 stopped,   0 zombie
Cpu(s): 0.0% us, 0.0% sy, 0.0% ni, 100.0% id, 0.0% wa, 0.0% hi, 0.0% si
Mem: 33276972k total, 29475484k used, 3801488k free,   317248k buffers
Swap: 8385880k total,        0k used, 8385880k free, 28772852k cached
PID USER      PR NI %CPU    TIME+ %MEM VIRT RES SHR S COMMAND
    1 root      16   0    0   0:02.88 0.0 2988 548 468 S init
    2 root      RT   0    0   0:00.18 0.0     0    0    0 S migration/0
    3 root      34 19    0   0:00.04 0.0     0    0    0 S ksoftirqd/0
    4 root      RT   0    0   0:00.11 0.0     0    0    0 S migration/1
    5 root      34 19    0   0:00.04 0.0     0    0    0 S ksoftirqd/1
    6 root      RT   0    0   0:00.09 0.0     0    0    0 S migration/2
    7 root      34 19    0   0:00.02 0.0     0    0    0 S ksoftirqd/2
    8 root      RT   0    0   0:00.07 0.0     0    0    0 S migration/3
    9 root      34 19    0   0:00.07 0.0     0    0    0 S ksoftirqd/3
   10 root       5 -10    0   0:00.00 0.0     0    0    0 S events/0
   11 root       5 -10    0   0:00.00 0.0     0    0    0 S events/1
   12 root       5 -10    0   0:00.01 0.0     0    0    0 S events/2
   13 root       5 -10    0   0:00.00 0.0     0    0    0 S events/3
   14 root      11 -10    0   0:00.00 0.0     0    0    0 S khelper
   15 root      15 -10    0   0:00.00 0.0     0    0    0 S kacpid
   64 root       5 -10    0   0:00.00 0.0     0    0    0 S kblockd/0
   65 root       5 -10    0   0:00.00 0.0     0    0    0 S kblockd/1
Any suggestions why 28gb memory is being used, I know this is not related to oracle but any suggestion will be a great help.

SGA Size

I will preface this by saying that this is probably a stupid question. I am pretty much an Oracle newbie, but I'm constantly trying to learn stuff. I have done a lot of reading on this topic, so I definitely have my own opinion, but everybody in my office thinks I'm crazy. So here goes...
We run a data warehouse in my shop on a server that has 6 CPUs and 12 GB of RAM, using Oracle 10.2.0.3. I was perusing our server's settings, and I noticed that the SGA was sized at 528 MB. It was not explicitly set in the parameter file, so I'm guessing that this value is calculated from the underlying pool sizes. Here's my stupid question: Isn't this not a good size for the SGA, given how much RAM we have?
Since we are a data warehouse, we should probably give ourselves a lot of RAM for the PGA's (according to Oracle's documentation). So I would like to crank up the SGA_TARGET to (at least) 4 GB and set the underlying pool sizes to 0 to let Oracle manage it all. But it's the fight of my life in my office. I am convinced that giving the SGA more RAM will increase performance.
So, am I crazy or not?

Sam, if everyone else in the office thinks you are crazy then you probably are; however, that does not mean that you are not correct in your argument.
The total amount of memory on the machine is not really a good factor for setting your SGA size. It is a limitation. The SGA size should be set based on the application load. Nevertheless the odds are good that doubling the SGA would have a measurable benefit since your machine has so much spare resource available.
Here is an idea. Take a seriies of statspack snapshots or use the AWR reports to measure the system during normal load. Then bounce the database and double the SGA. Rerun the same load and take the measurement. Compare. You ought to be able to sell this idea. When you see the benefit you can then repeat the procedure a couple of times to get up to a 2G, 3G, or 4G SGA.
Remember that your applicaiton will also need PGA and it is easy for the PGA requirements to equal or exceed the SGA requirements. At some point it may also be necessary to add another database instance (and SGA) to the machine so just because you have the memory does not mean that all of it should be put to immediate use.
HTH -- Mark D Powell --

SGA size is 1 GB free but oracle starts physical read/write,why ?

Similar Messages

Maybe you are looking for