MAX DB_WRITER_PROCESSES in 10g

Hi All,
I read the doc (Problem statement: [bug# 2441217]) from the metalink on the # of DBWRs. It says that Oracle changed the default # of DBWRs and MAX DBWRs since 9iR2.
Default: (CPU_COUNT + 7) / 8
Max: Min (CPU_COUNT / 2, 20)
Is it right? I tested this on HP
CPU_COUNT = 16, DB_WRITER_PROCESSES = 14
Based on the new algorithm MAX DBWRs are 8. But there are 14 DBWR processes up and running.
I think Oracle uses the privious aloghrithm even in 10g, which is based on dbblock_lru_latches.
Any idea?

Hi Sean,
As I understand it, db_block_lru_latches defaults to the following formula :
(CPU_COUNT x 6) or (DB_BLOCK_BUFFERS/50), whichever is less.
In earlier releases it defaulted to CPU_COUNT/2. If you set DB_BLOCK_LRU_LATCHES higher than this max value on systems where it is available (<9i) or set the undocumented parameter DBBLOCK_LRU_LATCHES higher, then Oracle just ignores this and sets it internally to the above calculated value. However it will be set to Num_pools x (CPU_COUNT/2) if multiple buffer pools (default, recycle, keep and any of the multi-block size pools) are setup.
More notes shere:
http://www.dba-oracle.com/oracle_tips_cpu_count_wrong.htm

Similar Messages

Max vs. 10g feature

I would like to have records with the max lfdnr. Does anybody know a new 10g-feature instead of my statement?
SQL> with dummy as ( select 25 Nr#, 1 LfdNr, to_date('30.06.1982','dd.mm.yyyy') end_date from dual union all
2                  select 25 Nr#, 2 LfdNr, to_date('30.06.1988','dd.mm.yyyy') end_date from dual union all
3                  select 25 Nr#, 3 LfdNr, to_date('30.11.1997','dd.mm.yyyy') end_date from dual union all
4                  select 25 Nr#, 4 LfdNr, to_date('31.12.2009','dd.mm.yyyy') end_date from dual union all
5                  select 44 Nr#, 1 LfdNr, to_date('07.05.2002','dd.mm.yyyy') end_date from dual union all
6                  select 44 Nr#, 2 LfdNr, to_date('01.08.2006','dd.mm.yyyy') end_date from dual union all
7                  select 44 Nr#, 3 LfdNr, null end_date from dual union all
8                  select 62 Nr#, 1 LfdNr, to_date('25.04.2006','dd.mm.yyyy') end_date from dual union all
9                  select 62 Nr#, 2 LfdNr, to_date('11.07.2009','dd.mm.yyyy') end_date from dual union all
10                  select 11 Nr#, 1 LfdNr, to_date('07.05.2002','dd.mm.yyyy') end_date from dual union all
11                  select 11 Nr#, 2 LfdNr, to_date('01.08.2006','dd.mm.yyyy') end_date from dual union all
12                  select 11 Nr#, 3 LfdNr, to_date('21.09.2007','dd.mm.yyyy') end_date from dual)
13 select Nr#, LfdNr, end_date
14 from dummy
15 where lfdnr IN (select max(lfdnr) from dummy
16           group by nr#)
17 and end_date >= sysdate
18 group by nr#, LfdNr, end_date
19 /
       NR#      LFDNR END_DATE
        25          4 31.12.09
        62          2 11.07.09
SQL>

I doubt your current query logic. You query only the LFDNR in the subquery but I think you should also include the NR# like that:
with dummy as ( select 25 Nr#, 1 LfdNr, to_date('30.06.1982','dd.mm.yyyy') end_date from dual union all
                select 25 Nr#, 2 LfdNr, to_date('30.06.1988','dd.mm.yyyy') end_date from dual union all
                select 25 Nr#, 3 LfdNr, to_date('30.11.1997','dd.mm.yyyy') end_date from dual union all
                select 25 Nr#, 4 LfdNr, to_date('31.12.2009','dd.mm.yyyy') end_date from dual union all
                select 44 Nr#, 1 LfdNr, to_date('07.05.2002','dd.mm.yyyy') end_date from dual union all
                select 44 Nr#, 2 LfdNr, to_date('01.08.2006','dd.mm.yyyy') end_date from dual union all
                select 44 Nr#, 3 LfdNr, null end_date from dual union all
                select 62 Nr#, 1 LfdNr, to_date('25.04.2006','dd.mm.yyyy') end_date from dual union all
                select 62 Nr#, 2 LfdNr, to_date('11.07.2009','dd.mm.yyyy') end_date from dual union all
                select 11 Nr#, 1 LfdNr, to_date('07.05.2002','dd.mm.yyyy') end_date from dual union all
                select 11 Nr#, 2 LfdNr, to_date('01.08.2006','dd.mm.yyyy') end_date from dual union all
                select 11 Nr#, 3 LfdNr, to_date('21.09.2007','dd.mm.yyyy') end_date from dual)
select Nr#, LfdNr, end_date
from dummy
where (nr#,lfdnr) IN (select nr#, max(lfdnr) from dummy
         group by nr#)
and end_date >= sysdate ;If you want to get rid of the subquery you don't need a 10g feature as already shown. If you want to use analytic functions: These are already available in 9i:
with dummy as ( select 25 Nr#, 1 LfdNr, to_date('30.06.1982','dd.mm.yyyy') end_date from dual union all
                select 25 Nr#, 2 LfdNr, to_date('30.06.1988','dd.mm.yyyy') end_date from dual union all
                select 25 Nr#, 3 LfdNr, to_date('30.11.1997','dd.mm.yyyy') end_date from dual union all
                select 25 Nr#, 4 LfdNr, to_date('31.12.2009','dd.mm.yyyy') end_date from dual union all
                select 44 Nr#, 1 LfdNr, to_date('07.05.2002','dd.mm.yyyy') end_date from dual union all
                select 44 Nr#, 2 LfdNr, to_date('01.08.2006','dd.mm.yyyy') end_date from dual union all
                select 44 Nr#, 3 LfdNr, null end_date from dual union all
                select 62 Nr#, 1 LfdNr, to_date('25.04.2006','dd.mm.yyyy') end_date from dual union all
                select 62 Nr#, 2 LfdNr, to_date('11.07.2009','dd.mm.yyyy') end_date from dual union all
                select 11 Nr#, 1 LfdNr, to_date('07.05.2002','dd.mm.yyyy') end_date from dual union all
                select 11 Nr#, 2 LfdNr, to_date('01.08.2006','dd.mm.yyyy') end_date from dual union all
                select 11 Nr#, 3 LfdNr, to_date('21.09.2007','dd.mm.yyyy') end_date from dual)
,dummy2 as (
select nr#, lfdnr, end_date, row_number() over (partition by nr# order by lfdnr desc) as row_no
from dummy)
select Nr#, LfdNr, end_date
from dummy2
where row_no = 1
and end_date >= sysdate ;Regards,
Randolf
Oracle related stuff:
http://oracle-randolf.blogspot.com/
SQLTools++ for Oracle:
http://www.sqltools-plusplus.org:7676/
http://sourceforge.net/projects/sqlt-pp/

Increase max datafiles in 10g

Hi All,
Is it necessary to recreate a control file in 10gR1 to increase maxdatafiles parameter.
Is it not enough if I just set db_files parameter to larger number?
Thanks,
Hal.

What do you say about this?
============================
maxdatafiles is set at 6 in my controlfile
select * from V$CONTROLFILE_RECORD_SECTION where type = 'DATAFILE';
TYPE RECORD_SIZE RECORDS_TOTAL RECORDS_USED FIRST_INDEX LAST_INDEX LAST_RECID
DATAFILE 180 6 6 0 0 2
select count(*) from dba_data_files;
COUNT(*)
6
show parameter db_files
NAME TYPE VALUE
db_files integer 8
create tablespace dummy03 datafile '/tmp/dummy03.dbf' size 1m
Tablespace created.
create tablespace dummy04 datafile '/tmp/dummy04.dbf' size 1m;Tablespace created.
I have reached 8 datafiles now, so the following shoudl fail
create tablespace dummy05 datafile '/tmp/dummy05.dbf' size 1m;
ERROR at line 1:
ORA-00059: maximum number of DB_FILES exceeded
and it failed!
and if I check my alert I actually see this:
kccrsz: expanded controlfile section 6 from 7 to 8 records
number of logical blocks in section remains at 1
kccrsz: expanded controlfile section 4 from 6 to 7 records
number of logical blocks in section remains at 1
Completed: create tablespace dummy03 datafile '/tmp/dummy03.
Mon May 9 19:09:39 2005
create tablespace dummy04 datafile '/tmp/dummy04.dbf' size 1m
Mon May 9 19:09:39 2005
kccrsz: expanded controlfile section 6 from 8 to 9 records
number of logical blocks in section remains at 1
kccrsz: expanded controlfile section 4 from 7 to 8 records
number of logical blocks in section remains at 1
kccrsz: expanded controlfile section 6 from 9 to 10 records
number of logical blocks in section remains at 1
DB_FILES[8] exceeded (fno=9 kcfdpk=8)
ORA-59 signalled during: create tablespace dummy05 datafile '/tmp/dummy05....
now I will increase only db_files parameters from 8 to 10
show parameters db_files
NAME TYPE VALUE
db_files integer 10
create tablespace dummy05 datafile '/tmp/dummy05.dbf' size 1m;
Tablespace created.
wow I didnt recreate controlfile and I am able to create new datafiles just by increasing db_files!
create tablespace dummy06 datafile '/tmp/dummy06.dbf' size 1m;
Tablespace created.
next should fail because I reached 10
create tablespace dummy07 datafile '/tmp/dummy07.dbf' size 1m;
create tablespace dummy07 datafile '/tmp/dummy07.dbf' size 1m
ERROR at line 1:
ORA-00059: maximum number of DB_FILES exceeded
and this is 8.1.7.4
check alert again
create tablespace dummy05 datafile '/tmp/dummy05.dbf' size 1m
Mon May 9 19:12:21 2005
kccrsz: expanded controlfile section 4 from 8 to 9 records
number of logical blocks in section remains at 1
Completed: create tablespace dummy05 datafile '/tmp/dummy05.
Mon May 9 19:13:08 2005
create tablespace dummy06 datafile '/tmp/dummy06.dbf' size 1m
Mon May 9 19:13:08 2005
kccrsz: expanded controlfile section 6 from 10 to 11 records
number of logical blocks in section remains at 1
kccrsz: expanded controlfile section 4 from 9 to 10 records
number of logical blocks in section remains at 1
Completed: create tablespace dummy06 datafile '/tmp/dummy06.d
Mon May 9 19:13:24 2005
create tablespace dummy07 datafile '/tmp/dummy07.dbf' size 1m
Mon May 9 19:13:24 2005
kccrsz: expanded controlfile section 6 from 11 to 12 records
number of logical blocks in section remains at 1
DB_FILES[8] exceeded (fno=11 kcfdpk=10)
ORA-59 signalled during: create tablespace dummy07 datafile '/tmp/dummy07.d...
check maxdatafiles again and it´s magically set to 10!
select * from V$CONTROLFILE_RECORD_SECTION where type = 'DATAFILE';
TYPE RECORD_SIZE RECORDS_TOTAL RECORDS_USED FIRST_INDEX LAST_INDEX LAST_RECID
DATAFILE 180 10 10 0 0 6
dump the controlfile and it´s 10 too!
STARTUP NOMOUNT
CREATE CONTROLFILE REUSE DATABASE "LNX817" NORESETLOGS ARCHIVELOG
MAXLOGFILES 16
MAXLOGMEMBERS 3
MAXDATAFILES 10
MAXINSTANCES 8
MAXLOGHISTORY 226

Question to load the entire database into memory.

I am planing to load the whole database into memory. Suppose mydb is 10G. Then I plan Max Memory for 10G. Then I can create a named cache with 10G and bind the mydb to this cache. Is this the best way to load entire db into memory?
If the whole db can be loaded into memory, if procedure cache, cache for tempdb and all other params are not important any more? Or still need to follow common practice to configure memory params?

Hi Kent,
12-15GB sounds reasonable.
I recommend always including your version with your initial posting (unless the version simply doesn't apply to the question). Particularly when running an unusual version, and 12.5.x has been end-of-lifed long enough to be unusual now. Are you running SAP Applications on this system? If not, please questions post to the SAP Adaptive Server Enterprise (SAP ASE) for Custom Applications space instead, it is the space for general-purpose ASE questions.
Cheers,
-bret

Setting up automatic reminder emails

I have regular (weekly, biweekly, monthly) conference calls and wish to send the invitees an automatic reminder email 1 day before and 1 hour before each conference call. How do I set that up in iCal?

Hi ,
To have the failover from standby to have automatically should have the following
1) Observer in place
2) The DG configuration is in either max availabilty or max performace ( in 11g) or max availability in 10g
3) You should have enabled flash back enabled on both primary and standby
4) You should set failover threshold property in broker
5) To have transparent application failover from the client , the client tns should have the address for both primary and standby

Question on FXO and CCM

If i have 8 POTS lines connected to an FXO card in a router, and I want to have an IP phone have call waiting, how would I set this up?
My thought process was this:
Config (2) 1mbs from the local LEC to have these in a hunt group. Basically one number available to (2) analogs. First line is busy, LEC rolls it to the second line, callmanager is configured to pick that number and ring the IP Phone.
Will this work? Of course in a PRI, this is fine, but is there any documentation for this setup with CallManager, 7600 router with FXO card.
Thanks much!

Hi,
"However, in 10g i noticed even though we bounce the DB numerous times....the size of UNDO and TEMP remains the size.
This leaves the only option of either increasing their size depending on the needs or
by dropping and re-creating those TS's (which i would think not suggestable on Production)"
What is reason to recreate ? If it is not suggested on production why it will work on test or development?
"Any suggestions on how to retain the sizes after they reach their max sizes in 10g?"
I suggest to identify the requirement of your UNDO and TEMP tablespace and retain it, I do not encourage autoextend or dynamic growth of UNDO/TEMP tablespaces.
Also pl note, UNDO tablespace after growing will always retain the space, it will not resize automatically.
Dilipkumar Patel.

Question on 9140 and 9216

Hi All,
One of our customer is running Cisco MDS 9140 switch and we have to test the switch with out storage system.
To test their configuration we don't have that switch in our lab so I am trying to find out which other switch has the hardware compatibility with 9140.
I tried to look for a document which could fetch us that information. Can anyone help me here? Which other switch uses the same ASIC and data path like 9140?
Thanks,
Malhar

Hi,
"However, in 10g i noticed even though we bounce the DB numerous times....the size of UNDO and TEMP remains the size.
This leaves the only option of either increasing their size depending on the needs or
by dropping and re-creating those TS's (which i would think not suggestable on Production)"
What is reason to recreate ? If it is not suggested on production why it will work on test or development?
"Any suggestions on how to retain the sizes after they reach their max sizes in 10g?"
I suggest to identify the requirement of your UNDO and TEMP tablespace and retain it, I do not encourage autoextend or dynamic growth of UNDO/TEMP tablespaces.
Also pl note, UNDO tablespace after growing will always retain the space, it will not resize automatically.
Dilipkumar Patel.

Olioweb QOS error and Test Run is not compliant

Hi,
I made 3 tile setup and test is failing for Olioweb QOS. When I run first 2 tile, test runs fine and run is compliant. Later I ran oly 3rd tile and test run is compliant without any issue.
When I start all 3 tiles together I get olioWeb QOS error for first 2 tiles. Sometimes all 3 tiles gives OlioWeb QOS error.
Attached the result.
Please look into it and let me know if anthing need to be changed in setup.
Thanks,
Suresh

jpschnee,
I have run only olioweb workload for all the three tiles and I do not see any failure in Olioweb.wrf. I have attached the result for the run. It looks like clean run when only Olioweb load is ran.
Again I ran the 3 tile for all workloads . I observed QOS error happens for all three tiles. i.e olioweb workloads. previosuly only for 2 tiles it gave QOS error.
I am trying to find the subsystem which is saturating in my setup. As I asked I am expecting some alarms from Vcenter when resources peaks to its max.
I am seeing CPU is taking less than 70% in all ESXi Servers. Memory less than 20%.
Disk Max latency is less than 20ms.
Network is max 500Mbps for 10G network connectivity.
Do you see above resource utilization problematic?>
Thanks,
Suresh

Cj44 and "test run"

Hi ABAP experts,
I'm calling CJ44 Tcode in my program via "call transaction". Afterwards I doing my own other things.
How can I get the value of "Test run" checkbox so I can run "my own things" only if the real run is performed?
Any idea, suggestion?
Thanks

jpschnee,
I have run only olioweb workload for all the three tiles and I do not see any failure in Olioweb.wrf. I have attached the result for the run. It looks like clean run when only Olioweb load is ran.
Again I ran the 3 tile for all workloads . I observed QOS error happens for all three tiles. i.e olioweb workloads. previosuly only for 2 tiles it gave QOS error.
I am trying to find the subsystem which is saturating in my setup. As I asked I am expecting some alarms from Vcenter when resources peaks to its max.
I am seeing CPU is taking less than 70% in all ESXi Servers. Memory less than 20%.
Disk Max latency is less than 20ms.
Network is max 500Mbps for 10G network connectivity.
Do you see above resource utilization problematic?>
Thanks,
Suresh

Max size of integer in oracle 10g

Hi,
What will be the max size of integer in oracle 10g
Any idea??
Thanks

Read exactly what others have written about integer representation in Oracle and read the rest of the documentation in the Oracle manual about data types and numbers. That is the only way you will learn what you need.
Your code does not work, because you have written it wrong! That is why you need to read the manuals properly and understand them. Then you will know what you should have written, and why your code does not work.
Hi,
its said that the maximum precision for a number is 38 but the below code does execute....
declare
a number;
begin
a :=
111111111111111111111111111111111111111111111111111111
111111111111111111111111111111111111111111111111111111
111111111111111111;
end;
The same if i declare a as integer will fail during execution
Any ideas plz
ThanksA number() is a floating point number. An integer is a fixed point number. There is a difference. Try number (38) as your data type.
By my count you have 126 1's in your number. You have already been told that Oracle stores integers with up to 38 digits. No wonder it fails. Lose about 90 of the 1's and it might work.
John

In Oracle 10g, what is the max number of values we can pass to 'IN' Clause

Hi,
In my SQL query i have a scenario like, i have to pass the values to 'IN' Clause dynamically from other application. I want to know what is the the max number of values i can pass to 'IN' Clause.
i have searched in many pages i seeing answers like 'may be 1000' , 'not sure'....
Could any one tell me the exact limit of IN Clause in 10g.
Thanks in Advance.
Chandran

Hi, Chandran,
Chandran M wrote:
In my SQL query i have a scenario like, i have to pass the values to 'IN' Clause dynamically from other application. I want to know what is the the max number of values i can pass to 'IN' Clause.
i have searched in many pages i seeing answers like 'may be 1000' , 'not sure'....Don't overlook [Dont want to know|http://forums.oracle.com/forums/thread.jspa?messageID=2683385&#2683385]. This may explain why people are reluctant to give a straight answer.
Could any one tell me the exact limit of IN Clause in 10g.1000 items such as IN (1, 2, 3, ..., 1000)
No limit on the number of sets: IN ((1), (2), (3), ...) *<=== NOT QUITE!* See below
No limit in sub-queries: IN (SELECT ...)
Edited by: Frank Kulash on Jun 3, 2009 9:43 AM
The limit of 1000 applies to sets of single items: (x) IN ((1), (2), (3), ...)
There is no limit if the sets contain two or more items: (x, 0) IN ((1,0), (2,0), (3,0), ...)
Thanks to Michaels for pointing this out.
Edited by: Frank Kulash on Jun 3, 2009 10:22 AM
Fixed link

Max connections in Oracle 10g Express

Hi,
How much connections are possible in Oracle 10g Express ?

I do the following test:
I format the test PC, install the O.S. and install Oracle 10g Express and run the following code:
package testedb;
import java.sql.Connection;
import java.sql.DriverManager;
import java.sql.SQLException;
import java.util.ArrayList;
import java.util.List;
public class TesteDB {
     * @param args
     public static void main(String[] args) {
          List<Connection> connections = new ArrayList<Connection>();
          try
     Class.forName ("oracle.jdbc.driver.OracleDriver");
     while(true)
          connections.add(DriverManager.getConnection("jdbc:oracle:thin:@albatroz:1521:XE", "system", "xxx"));
          //connections.add(DriverManager.getConnection("jdbc:oracle:thin:@albatroz:1521:MYDB", "system", "xxx"));
          System.out.println("connections: "+connections.size());
     catch (ClassNotFoundException e)
     System.out.println ("Could not load the driver");
     e.printStackTrace ();
     } catch (SQLException e) {
          System.out.println("----------------------------------------------------------");
               System.out.println("Qtd of connections: "+connections.size());
               System.out.println("----------------------------------------------------------");
               e.printStackTrace();
After the first test I format again the PC, install the O.S. and install Oracle 10g Standard and run the same code, changing only the database name.
Well, I get the following results:
- Oracle 10g Express: I get 19 connections;
- Oracle 10g Standard: I get 181 connection;
Final report:
With the same hardware (1GB RAM) and O.S. (Windows XP Professional) test, the Oracle 10g Express has any restrictions at the question max connections.
Message was edited by: Hugo Haas
Hugo Haas

Implement MAX / JOIN in ODI 10g?

What is the approach for using MAX functions in ODI 10g? I need it to filter source data:
SELECT SRC_TAB.*
FROM SRC_TAB
INNER JOIN
(SELECT MAX(COL1) COL1, COL2 FROM SRC_TAB GROUP BY COL2) B
ON SRC_TAB.COL1=B.COL1 AND SRC_TAB.COL2 = B.COL2
Luckily this issue has been addressed in ODI 11g.
Thank you.

That's not good especially when you have large volume of data.
In this case, in 10g it's better to rely on a underline view rather than 2 interfaces.

Max phisical size in forms 10g?

i need to know what is the max fisical size in a form 10g? 1Kb, 1Mb ?

There are whole lot of theoretical limits that are too numerous to list here and too theoretical to really matter but the upshot is that you can have pretty big fmx file sizes if you really want to. Using several smaller forms is probably a better overall strategy since memory can be managed better.

10g db_writer_process

Hi Brother,
i am using 10g on T2000 with 16core, would i increase the db_writer_process? default is 2 and would like to change to 8 or 16. pls advise
Thanks

take a look at these metalink doc...
Subject:      Understanding and Tuning Buffer Cache and DBWR
     Doc ID:      62172.1
Subject:      DBWR in Oracle8i
     Doc ID:      105518.1

MAX DB_WRITER_PROCESSES in 10g

Similar Messages

Maybe you are looking for