RBS tablespace abnormal growth

I am checking the tablespace size , i am seeing growht of RBS tablespace / it increases 297 MB in one day...Can any body tell me what is the reason & how i can solve that?

As you wish. You can put the other datafile to another disk. Oracle will stripe the data auomatically. Or you can enlarge your existing datafile if you don't have another disk to use.
Less datafiles means less administration effort. So you should trade off if it's worth to add a second or third datafile to your TEMP Tablespace. But more performance will perhaps also increase your salary ;-)
Dim

Similar Messages

Oracle Database abnormal growth

Dear all,
Can anybody suggest me, what is the best method to analyze and check the abnormal growth of database?
Currently our database size is 3.8 Tera , we have more then 1300 concurrent users and we are collecting statistics info on daily basis from DB02, ST04, ST10, and program RSSTAT10.
Can any body suggest me an answer with regards to various options available?
Thanks and regards
Asif
Message was edited by: Mohammed Asif
Message was edited by: Mohammed Asif

Mohammed,
Lot's of work to do. No easy fix for that.
As far as GLPCA you need to work with application teams to do data archiving. As far as I know, this is the only way you can shrunk this table.
TST* tables are spool/temse related tables. You can go over OSS notes 160083 48400 and make sure that spool and temse consistence check jobs are running periodically based on your company's retention period limitations.
CDCLS (data table) is change pointer related. You can get more maeningful information on CDHDR ( header table) such as how old data is out there. You need change pointer deletion/reorganization or archiving jobs scheduled. I do not have OSS notes handy but check OSS and schedule changepointer deletion jobs based on your company's retention period requirements.
After running cleanup jobs, do not forget to run an analyze table statement at the DB level and see how much space you have gained. If it is more than 40%, I recommend full table reorganization otherwise unbalanced indexes might cause you performance loss and you cannot allocate the space back to tablespace as freespace to be shared by other tables.
Good luck!
Neval

RBS Tablespace Issue

Hi All,
Greetings,
Oracle 8.0.6
SunSolaris
My Rollbacksegment Tablespace datafile size is 1640 MB
but the actual used space is 217 MB.
is there any way to compress the size of 1640MB
Please Help me in this
Thanks in Advance
Adil

Hi,
01628, 00000, "max # extents (%s) reached for rollback segment %s"
// *Cause: An attempt was made to extend a rollback segment that was
// already at the MAXEXTENTS value.
// *Action: If the value of the MAXEXTENTS storage parameter is less than
// the maximum allowed by the system, raise this value. => Alter your RBS and allow more than 121 extents to be created in this RBS.
Also i found that RBS tablespace is marked as PERMANENT.Yes, it's normal.
..it means the rollback data does not get flushed out periodically?No
Also Should this tablespace be temporary? No, this can't be
How can i counter this problem?Free advise: RTFM about Rollback Segments! Start here (8i doc) or read the Concepts book.
Regards,
Yoann.

Can size of RBS tablespace be reduced?

Hi folks.
Hope this is the right forum for this question.
Anyway, we're running Oracle 8.1.6 Standard Edition on Win NT
4.0. A few months ago, we ran a SQL Update statement on a table
with 176,000 rows and then did a Commit. As a result our RBS
(rollback?) tablespace grew quite large. Also, as a result, we
now do more frequent Commits. Anyway, my question is: can we
reduce the size of the RBS tablespace and if so how? I tried to
do it using DBA Studio but it wouldn't let me. Do I first have
to remove some Rollback Segments? (We have 6 now, each one 8,192
KB in size.)
Any hints, suggestions, etc. would be appreciated.
Thanks.

Hello Craig
To reduce the size of rollback segment you may issue the
following statement :
alter rollback segment <Segment-Name> shrink to <integer> K ;
If you have given the Optimal value for the specified Rollback
segment, you can shrink it to Optimal value as :
alter rollback segment <Segment-Name> shrink ;
Hope it will help you.
Manish Gupta

Abnormal growth of tablespace

Hi all,
How to know the growth of a tablespace of last one hour.
Exp:while monitoring a particular filesystem it showed me 48.896 and i updated that.
After 10 min when my lead is checking 48.696,he is saying that there cant be 200MB growth in 10 min.
How to find out.........plz

Without knowing more about your configuration it's almost impossible to help you.
What's the autoextend size of the tablespaces, that's affected? If this is 200 MB then it's correct, the database increased the space for 200 MB but they are not yet occupied by data.
But without knowing more about your Oracle configuration in detail, it's impossible to dig that down.
Markus

How to coalesce RBS tablespace??

Hi,
I´m using a database 9.2.0.7 that is using RBS and not UNDO Tablespace.
This tablespace is full and i can´t empty this.
How can i do that? Empty this tablespace. No one is logged on this database, so this tablespace don´t have any transaction on this.
ORA-01562: failed to extend rollback segment number 34
ORA-01650: unable to extend rollback segment RBS1 by 128 in tablespace RBS
Tks,
Paulo.

Cross check your 10g init.ora parameter with Metalink doc id 216205.1 (This note/doc mentioned lots of parameters for performance) and make sure you have collected stats for 10g database. If you are using global temporary table in 10g then delete stats for global temporary table and lock stats like
ONLY for GLOBAL Temporary Table
analyze table scott.GTEMP delete statistics;
execute DBMS_STATS.Lock_Table_Stats(ownname=>'SCOTT', tabname=>'GTEMP');
Hope performance will improve drastically.
Regards,
Virag Sharma

Dbacockpit (tablespace statistics, growth/day)

We use new dbacockpit according to note 1028624 and have implemented the corrections of note 1042725, 1046668, 1052909, 1076147, 1081265, 1066044.
In difference to the old DB02 (DB02OLD) , I don't see detailed information about the tablespace growth per day:
In new DB02: tablespaces -> Overview, Summary History I see only one line per tablespace with
Begin date End date   Tablespace name   Size(MB) Chg.size/day Free(MB) Chg.free/day Used(%) Chg.Used/day(%/day) Autoextend Total size(MB), etc.....
but I can't see detailed growth for EACH day like in DB02OLD, when double-clicking on the specific tablespace:
DB02OLD:
29.08.2007             4.608.000 1462.272    1.398.240   3.209.760    138.240     69     28-    242        0        910        12
28.08.2007             3.145.728        0       74.208   3.071.520    927.120     97     29     242        0        898        39
27.08.2007             3.145.728 1048.576    1.001.328   2.144.400    102.200     68     29-    242        0        859         6
26.08.2007             2.097.152        0       54.952   2.042.200    186.880     97      9     242        0        853        13
Is this normally ? Sometimes it's very important to know , when the growth exactly took place and not see the average growth per day....
Kind regards,
Uta

Thanks a lot. The handling of the new db02 is not very intuitive...
Unfortunately the new DB02 doesn't supply the same functions than the old db02.
I hope, that the functionality of the new DB02 will be better with the next supportpackages....
I see following disadvantages:
1) you have to SELECT the tablespaces explicitly (also * possible ) to see history data
2) no sort of tablespaces by changes/month possible, because thes data is only on the detail screen and history tab available, but not in the overview of selected tablespaces
3) if different tablespaces selected, the last selected tab is not memorized. To see the detailed history for several tablespaces, you have to click 3 times (name of tablespace, history tab, months tab)
Perhaps the developpers also read this forum.....;-)
Regards,
Uta

Abnormal growth of Data file

If i import a dmp file of size 20MB, it is creating a data file of size 300MB or more than that (in one case 2 GB). Pl. tell me the appropriate storage parameters for not to increase the data file size abnormally.
thank u
[email protected]

Hi Angel
In that case passed 2 possible things:
1.- The transfer to the tape or from the tape ot disk was not in binary mode
2.- The export utility did not finish succesful the exportation when somebody was doing that activity.
Joel PÃ©rez
http://www.oracle.com/technology/experts

Tablespace history, growth, which tables, tx DB02

Hi,
Here is the scenario : I want to run a unix script (with sql commands) which selects information from the "history tables" (transaction DB02 - History) and then creates a flat file which can be emailed.
In which tables are the history information saved ?
Set up : BW 3.5/NW04, DB2 v8, AIX 5.2.
Hints and help will be appreciated.
Thanks in advance.
Christian

Hi,
direct evaluation of these tables is a little bit tricky. There are several tables, all starting with the prefix DB6PMH. And there is mostly a header table with suffix HD and a detail table. E.g. DB6PMHT and DB6PMHT_HD contain table and index history, DB6PMHXP_HD and DB6PMHXP contain tablespace history. For the details on how to interpret the columns you can look into the ABAP dictionary, mostly there is a short documentation there. But you should be aware of the fact, that this is not an 'official' interface and the layout of these tables might change in the future
Best regards
Ralf

Abnormal growth of library cache

I am using Oracle 10g. For the last few days I am getting ORA-04031 error frequently. Actually the library cache is growing much faster than usual. It is taking 20 minutes to become 300M from 12 M(when I restart the DB).
What may be the problem? Can anybody help me??

You can check if you have a problem with using bind variables using the following script, which I copied from AskTom:
create table magweg_lclc as select sql_text from v$sqlarea;
alter table magweg_lclc add sql_text_wo_constants varchar2(1000);
create or replace function remove_constants( p_query in varchar2 ) return varchar2
as
    l_query long;
    l_char varchar2(1);
    l_in_quotes boolean default FALSE;
begin
    for i in 1 .. length( p_query )
    loop
        l_char := substr(p_query,i,1);
        if ( l_char = '''' and l_in_quotes )
        then
            l_in_quotes := FALSE;
        elsif ( l_char = '''' and NOT l_in_quotes )
        then
            l_in_quotes := TRUE;
            l_query := l_query || '''#';
        end if;
        if ( NOT l_in_quotes ) then
            l_query := l_query || l_char;
        end if;
    end loop;
    l_query := translate( l_query, '0123456789', '@@@@@@@@@@' );
    for i in 0 .. 8 loop
        l_query := replace( l_query, lpad('@',10-i,'@'), '@' );
        l_query := replace( l_query, lpad(' ',10-i,' '), ' ' );
    end loop;
    return upper(l_query);
end;
update magweg_lclc set sql_text_wo_constants = remove_constants(sql_text)
select sql_text_wo_constants, count(*)
from magweg_lclc
group by sql_text_wo_constants
having count(*) > 1
order by 2
drop function remove_constants
drop table magweg_lclc
/All statements you see are not using bind variables. The count(*) column determines how much it is bugging you. The only thing left to do, is to spot the queries in the code and talk some sense in the developers responsible for that code ;-)
Hope this helps.
Regards,
Rob.

Growth tablespace PSAPBWPODS

Hi Gurus, a few months ago we was create a Z process chain that is execute daily for improve performance. The PC do this 5 steps for some BPS infocubes:
1.     Load Mode On
2.     Create DB Statistics
3.     Delete DB Indexes
4.     Create Indexes
5.     Plan Mode On
The problem is that our tablespace PSAPBWPODS is full. I checked that since we are running this process chain, the tablespace was growth 2.000 Mb per month, but before of this was only 200 Mb per month.
How is possible to clean up this tablespace? I think that stadistics are the cause of this problem, is possible to delete it?
Thanks.

Hi,
You can investigate when and why the tablespace grew with DB02OLD > Space Statistics for tablespace and then checking growing per week, month. And also use ST04 > Space > Tablespace > Detailed Analysis to check what happened exactly.
Br,
Javier

Managing Undo tablespace & roll back segments

Hi,
I have oracle 9.1 version and have created 5g undo tablespace apart from 20m rbs tablespace for 15 rollback segments. But it is seen that whenever i give UNDO_MANAGEMENT=AUTO then rollback segments are created using undo tablespace and not allowing rollback segment in rbs tablespace ONLINE. In the reverse happens whenever I give UNDO_MANAGEMENT=MANUAL , the rollback segments in the rbs tablespace gets focus and roll back segments in the Undo tablespace goes offline. Is there anything that I can do so that both undo tablespace and redo segments in the rbs tablespace gets working together so as to minimize the undo contention. Moreover, any query given takes long time when select query given with || in the subquery.
Please suggest the tuning need to do?

brain_heart wrote:
Hi,
I have oracle 9.1 version and have created 5g undo tablespace apart from 20m rbs tablespace for 15 rollback segments. But it is seen that whenever i give UNDO_MANAGEMENT=AUTO then rollback segments are created using undo tablespace and not allowing rollback segment in rbs tablespace ONLINE. In the reverse happens whenever I give UNDO_MANAGEMENT=MANUAL , the rollback segments in the rbs tablespace gets focus and roll back segments in the Undo tablespace goes offline. Is there anything that I can do so that both undo tablespace and redo segments in the rbs tablespace gets working together so as to minimize the undo contention.No that's the way it would work, you can either use the automatic undo or the manual one.not both. And its better to stick with the automatic one IMHO. Have a read of this link to understand more about undo.
http://download.oracle.com/docs/cd/B10501_01/server.920/a96521/undo.htm#8888
HTH
Aman....

Usage of RBS

I used database backup and do the restoration on a testing machine and I can
observe the rbs tablespace's space was used up after applying the archivelogs
to the accident time. However, there are no txn information inside (that's expected).My question is: can I, during the archivelogs being applied, to check the rollbacksegments utilization (e.g. by querying the v$transaction view)? Actually i tried this,but no rows were selected.

restore means recovery amongst others.
No ordinary transactions, so no info in v$transaction.
select * from dict
where table_name like '%RECOVERY%'
will show you which views provide info on recovery.
All data dictionary views are listed here.
Sybrand Bakker
Senior Oracle DBA

Growth PSAPBWPODS

Hi Gurus, a few months ago we was create a Z process chain in our SAP BI system that is execute daily for improve performance. The PC do this 5 steps for some BPS infocubes:
1. Load Mode On
2. Create DB Statistics
3. Delete DB Indexes
4. Create Indexes
5. Plan Mode On
The problem is that our tablespace PSAPBWPODS is full. I checked that since we are running this process chain, the tablespace was growth 2.000 Mb per month, but before of this was only 200 Mb per month.
How is possible to clean up this tablespace? I think that stadistics are the cause of this problem, is possible to delete it?
Thanks.

Neo wrote:
> Hi,
>
> In DB02old query all tables and sort them by size.
>
> Have you got lot of these type tables /BIC/B0000<nr>
>
>
> If yes are these in you ODS tablespace ???
>
> And if you query /BIC/000<nr>* does this account for a lot of tables taking alot of space ?
>
> Mark
> -
>
> Yes Mark when I check in tablespace PSAPBWPODS, the tables that have more size are like /BIC/B0000* (table part and index part)... What´s that mean??
>
> Thanks.
> NEO
Hi neo,
I have had similar tablespace growth and queried this with the functional people as it looks like that PSA was not being deleted
What we found out was since upgrading from BW 3.5 to BI 7.0 earlier in the year some characteristics of the process chains has changed.
The problem that we overlooked in BI 7.0 was that change logs were not being deleted.
Have a word with your functional people to go over the process chains.
In my case after we deleted the change logs and reorganizing the /BIC/B000260 and other tables I got 200GB back!!
Mark

System tablespace conversion error

Hi all,
I was practising the 8i to 9i upgrade in lab server, when I was trying to convert the dictionary managed tablespace to locally managed tablespace its throughing the error, I dont know what is the reason and what I have to do to eliminate this error
I used this stmt to convert,
EXECUTE DBMS_SPACE_ADMIN.TABLESPACE_MIGRATE_TO_LOCAL('SYSTEM');
The error I got :
BEGIN DBMS_SPACE_ADMIN.TABLESPACE_MIGRATE_TO_LOCAL('SYSTEM'); END;
ERROR at line 1:
ORA-10641: Cannot find a rollback segment to bind to
ORA-06512: at "SYS.DBMS_SPACE_ADMIN", line 216
ORA-06512: at line 1
I have removed the rbs datafile after dropping the rbs tablespace. and created undo. but the error still comes, dont know what to do
please help me out.
Thanks in advance :)

hi guys I have bounced the dba and restarted it, it worked fine, thanks

RBS tablespace abnormal growth

Similar Messages

Maybe you are looking for