Database Performance (Tablespaces and Datafiles)

Hi guys!
What's the best for performance in database, tablespace with various datafiles distribuited in diferents filesystems or tablespaces with various datafiles in only one filesystem?
Thanks,
Augusto

It depends on contents of the tablespaces, tablespace level LOGGING/NOLOGGING and env such as either OLTP or OLAP and LUN presentation to the server with RAID or without RAID and SAN Read per second and write per second.
In general, tablespace with various datafiles distribuited in diferents filesystems/LUN's is in practice for non dev/system test databases.
Moreover using ASM is better then standard filesystems.
Regards,
Kamalesh

Similar Messages

Tablespace and datafiles

I have 3 test databases with Oracle 11.2.3 on the same box
data.dbf and index.dbf were deleted from 2 of the databases in SQL and now EP Manager can't get to any of the tablespaces. when clicking on Server - Tablespace i get the following error:
java.sql.SQLException: ORA-01116: error in opening database file 2 ORA-01110: data file 2: '/u02/oradata/LAWTEST/dev_data01.dbf' ORA-27041: unable to open file IBM AIX RISC System/6000 Error: 2: No such file or directory Additional information: 3 Additional information: 4 Additional information: 4194304
I have tried dropping the tablespaces and refreshing the database from a Prod backup but no luck and have been searching the web for any info.
Any help would be greatly appreciated.

Hi,
I would like to ask where is stored in the database
the information regarding the tablespace and
datafiles (for example for a tablespace which
datafiles belong?). Is it in controlfile? I think no.
Where is it?
Thank you,
MihaelaHi,
You may query DBA_TABLESPACES or V$TABLESPACE to views to gain information on tablespaces. Similarly, DBA_DATA_FILES and V$DATAFILE views will list you the details related to the data files like tablespace name/id, file location ......etc..
Regards

Tablespace and Datafiles... HELP

Hi, I3m new to Oracle, I3m trying to install Oracle 8.1.5 under
Tru64 UNIX, I3m having a hard time creating the Tablespace and
Datafiles to the oracle User... I cannot find documentation
about it, can somebody here tell me how to do it????, it3s
urgent, Please HELP mee!!!!
null

Hi Ruben,
I'm an Ora admin working under NT.
Use oracle server manager to create datafiles & tablespaces.
You can find the executable file name for oracle server manager
(svrmgr??.exe for NT) under BIN irectory of ORACLE HOME.
Hope this will be tip for you.
Regards
Sukumar
Ruben Gomez (guest) wrote:
: Hi, I3m new to Oracle, I3m trying to install Oracle 8.1.5 under
: Tru64 UNIX, I3m having a hard time creating the Tablespace and
: Datafiles to the oracle User... I cannot find documentation
: about it, can somebody here tell me how to do it????, it3s
: urgent, Please HELP mee!!!!
null

Drop a tablespace and datafile when database is mounted

I tried to import data in my oracle 11g. It failed midway. I presumed that i did not size the redolog properly. Now my database cannot go beyound the mounted mode.
ORA-01157: cannot identify/lock data file 7 – see DBWR trace file
ORA-01110: data file 7: ‘ /data1/owprojects/xxxxx.dbf ‘
Its a new oracle installation and i do not need the corrupt datafile.
I cant take the tablespace offline neither can i drop the tablespace including its contents.
Do i drop from the Linux OS. If yes which other file to i edit to remove the pointer to this tablespace.
Thanks and Best Regards
J

Try:
alter database datafile 7 offline for drop;Example:
SQL> startup mount;
ORACLE instance started.
Total System Global Area 1071333376 bytes
Fixed Size                  1388352 bytes
Variable Size             658505920 bytes
Database Buffers          406847488 bytes
Redo Buffers                4591616 bytes
Database mounted.
SQL>
SQL> alter database open;
alter database open
ERROR at line 1:
ORA-01157: cannot identify/lock data file 5 - see DBWR trace file
ORA-01110: data file 5: 'C:\TEMP\TEST01.DBF'
SQL> alter database datafile 5 offline for drop;
Database altered.
SQL> alter database open;
Database altered.
SQL> archive log list;
Database log mode              Archive Mode
Automatic archival             Enabled
Archive destination            USE_DB_RECOVERY_FILE_DEST
Oldest online log sequence     396
Next log sequence to archive   397
Current log sequence           397
SQL> alter tablespace test drop datafile 5;
alter tablespace test drop datafile 5
ERROR at line 1:
ORA-03261: the tablespace TEST has only one file
SQL> drop tablespace test;
Tablespace dropped.
SQL>Please read http://docs.oracle.com/cd/E11882_01/server.112/e25494/dfiles004.htm#ADMIN11428.
Edited by: P. Forstmann on 19 mars 2012 18:53

Tablespace and datafile hot backup

Hi,
I wanted to if there is a command to copy a datafile using sqlplus command. I making a script that will backup the database while its online.I made a script for the alter tablespace begin backup but I wanted to know how to copy the datafiles without using the OS command,instead I wanted to use a sql commands. Is there any script or procedure to do this one.
thanks
Best Regards,
Antok

RMAN is the best utility to do the backup and recover activities. It take care copy and paste itself. Bellow given script is also usefull for hot backup
spool hotbackup_PROD3.sql;
set feedback off;
set verify off;
set pagesize 2500
set linesize 132
set serveroutput on size 50000;
declare
cursor tablespace is select tablespace_name from dba_tablespaces where tablespace_name
in('RBS1',
'STORAGE',
'SYSADM_IDX2',
'SYSADM_IDX3',
'SYSADM_IDX4',
'SYSTEM');
cursor datafile(tsname varchar2 ) is select file_name from dba_data_files
where tablespace_name = tsname;
begin
dbms_output.put_line('spool hotbackup_PRODc.log;');
dbms_output.put_line('alter system switch logfile;');
dbms_output.put_line('select ''log''||sequence#,sysdate from v$log where status = ''CURRENT'';
for tsname in tablespace
loop
dbms_output.put_line('alter tablespace '||tsname.tablespace_name||' begin backup;');
dbms_output.put_line('prompt Starting Backup for '||tsname.tablespace_name||' Tablespace')
for dname in datafile(tsname.tablespace_name)
loop
dbms_output.put_line('prompt Copying File '||dname.file_name||'.....');
dbms_output.put_line('!cp '||dname.file_name||' /backup2/hotbak'||dname.file_name);
end loop;
dbms_output.put_line('alter tablespace '||tsname.tablespace_name||' end backup;');
dbms_output.put_line('prompt Finished backing up '||tsname.tablespace_name||' Tablespace')
dbms_output.put_line('alter system switch logfile;');
end loop;
dbms_output.put_line('spool off;');
dbms_output.put_line('!mailx -s "KL_PROD521_HOTBACKUP1" [email protected] < /prod1/HOTBKUP/hotbackup_PRODc.log');
dbms_output.put_line('exit');
end;
spool off;
@hotbackup_PROD3.sql
exit;

Use of backing up a single tablespace or datafile

Hello,
I am reading the RMAN manual and I'm quite familiar with backup up a tablespace or datafile, but I can find very few uses for that. Backup up a tablespace is useful for TPITR, but since that needs another instance its of very little use in most production environments. I think you cannot restore an old version of tablespace in a normal database, unless the tablespace has long been made read-only, in which case RMAN's optimizations will do the trick.
Even less use I can find for datafile backup. I have absolutely no idea what you can do with a single datafile.
Can you please clarify me on the uses of these RMAN features?
Thank you.

Dear Albi!
Think of the following scenario:
You have a very large (let's say 1 Terabyte) production database that is split into n tablespaces. This database is in archivelog mode and a full backup of the hole database would take more than 10 hours.
scenario end.
In such a scenario I think you will not take a full database backup very often. Therfore you can backup portions (tablespaces and datafiles) of your database. This will take less time then a hole DBbackup.
If you have to restore your DB then RMAN will take all the files it needs from all your backup. RMAN uses always the most actual version of your datafiles. After the restore RMAN will take the archivelogs to recover all datafiles to the most actual point in time. And that's the point why partial backups of your db are not only usefull for TSPITR.
I hope I could make clear what the operative point is with partial backups.
Yours sincerely
Florian W.

To recover a database with old and new datafiles.

Hi;
Please Give a Solution for the following problem.
First I shutdown the databae.
Copy all the redolog,control file and Datafile except one.
One datafile i miss at the time of copy.
Then I start the database and do some transactions.
At that time my system tablespace has been corrupted.
So i copy the old backup and overwrite with the actul file.
So one file backup is not there, that i forgot to take.
At the time of opening the database it shows that datafile version is higher than the controlfile.
Now what i did is came to mount stage and type the following command.
ALTER DATABASE DATAFILE <FILENAME> OFFLINE DROP;
After that i open the database.
Now the problem is that missing file contain very important data.
Is it possible to recover the data?
Please give me an immediate solution.

Is the database running in archive log mode? Do you have any export dump files for the data in question?

Can i use "Oracle Database 12c: Performance Management and Tuning " training for getting certification on "Oracle Database 11g: Performance Tuning 1Z0-054"

i have taken "Oracle Database 12c: Performance Management and Tuning new" training from oracle university. Now i would like to get certified on "Oracle Database 11g: Performance Tuning 1Z0-054" exam. Is it possible ?

I essentially endorse and refer you to Matthews' and John's post above.
I would differ with slightly with Matthew because my guess is you would often be able to use like for like 12c training for an 11g certification ( I believe there are precedents). BEFORE ANYONE ASKS THE OTHER WAY DOESN'T HAPPEN.
.... but totally concur with Matthew you would ill advised to procede on that basis without one of:
- This being advertised as possible on the website : e.g. https://education.oracle.com/pls/web_prod-plq-dad/db_pages.getpage?page_id=654&get_params=p_id:169 ... option 2 show courses.
- Confirmation from Brandye
- Confirmation from an Oracle Certification Support Web Ticket ( http://education.oracle.com/pls/eval-eddap-dcd/OU_SUPPORT_OCP.home?p_source=OCP )
... The more common (and in my opinion usually better) way would be get your 11g DBA OCP ( or higher first) and then take the 1z0-054. I am almost certain they will accept your 12c course for the 11g DBA OCP.
If you are choosing the route of not being a 11g (or 12c ) DBA OCP first but are on option 2 and relying on the course for certification then the issue is more in the balance and you are even more strongly advised to get confirmation before proceding (remember if the rules need to be changed for you only then any profit out of the exam is lost).
In general my understanding is Oracle would prefer to encourage people to train on the latest version of product that is available for training and will prefer to avoid restrictions which would cause you to train at a lower version. ( This is simply my guess at Oracle University Policy ... personal opinion only).
Having said all I have said I'd encourage you to go with the advice of the earlier two posts.

Tablespace and performance issue

Hi,
I have Tablespace X and tablespace Y.
I have seven transaction tables in X, which has image tables in Y. X tablespace is my Online tablespace and Y my archive tablespace and I use the archive tablespace to store old data.
I use the same algorithm, only wrapped up in two processes, with only the source and destination alternating. The code performs extremely well for moving data X->Y. And ridiculously slow for Y-> X.
I have just started doing this process of archiving, so the data volume in both the sets of tables are the almost the same.
Any clue on what I could check up for here ?? What should be the problem ??
The trace file shows the similar plans both sides, only there are wait times in process Y->X.
Thanks for your time on this.
Ashok.

ahhh ... you are talking now about "internal" (vs. "external"?) volume instances. This implies probably you have really to talk to the responsible dba, as Justin already proposed.
One small correction, you always can do an "insert /*+ append */ ...". As Oracle will write into new datablocks then, you only would loose the last allocated free blocks. But in the case of the offline tables, this is not a problem as you only add data anyway therefore you'll not loose a lot. And disk space is not really a concern anymore these days.
For the other way around, you only had to do an "alter table X move tablespace <online-tbs>" afterwards to clean up all the unused blocks.
For the truncating, may be you can think about it again. You have a processing going on there ie there is some grouping/selecting criteria. Probably one of these criterias is a partitioning candidate? And with a partitioned table you would be back in the truncating game.
For the "I'm not let to DDL on the objects from inside the code" seems a rather funny restriction, as this doesn't let you do a truncate and instead you need a lot of (probably unnecessary) undo-tablespace for a delete. But as some dba's don't really know, what the job of a good developer is, you can probably ask your dba for a packaged procedure, where this table/partition gets truncated.

Performance reports and measures for Online Trading Customer's databases

We have a client from online trading commodity domain. there databases are running in healthy stat.
We on daily basis share with them AWR report of time span 9:30AM to 10:30AM (peak trading hours).
can you suggest what all recommendadtions we can include in this AWR report or how we can study on daily basis this AWR report extracted to mention recommendation to customer.
apart from this what other performance reports we can share with customer which will make cilent view their database stat in terms of performace tunning and all.
We can share daily database checks and all but I am trying to have some good performance end reports which can be extract to share with client??
What performace reports you are sharing or can suggest in this regards??
Thanks friends in advance.

Hi
Ankit Ashok Aggarwal wrote:
We have a client from online trading commodity domain. there databases are running in healthy stat.
We on daily basis share with them AWR report of time span 9:30AM to 10:30AM (peak trading hours). I have written a few blog posts about AWR analysis (http://savvinov.com/category/awr/) -- some of them may be useful to you. However, AWR reports are not so good for monitoring purposes. Monitoring is about trends -- AWR doesn't have that. Plus, application performance should be measured in its native metrics (KPIs or "key performance indicator"). For instance, for an online shop that would be the number of orders processed, average time it takes to process an order etc. Low level stats such as the number of table scans per unit time or IO stats doesn't really tell whether or not an application is performing satisfactorily.
From the database point of view, a simple overview of the OEM Performance Page should be enough to get a basic idea whether the database is ok.
can you suggest what all recommendadtions we can include in this AWR report or how we can study on daily basis this AWR report extracted to mention recommendation to customer.I've seen a lot of such recommendations -- 99% of them are garbage, and the customers treat it accordingly. Unless there are clear signs of a performance issue, recommendations are generally neither necessary, nor possible (without additional information about the application). Ain't broken -- don't fix it.
apart from this what other performance reports we can share with customer which will make cilent view their database stat in terms of performace tunning and all.
We can share daily database checks and all but I am trying to have some good performance end reports which can be extract to share with client??Ideally, you should have SLAs with your customers which should clearly define how much time a certain user action should take. Without an SLA and/or specific complaints from the user there is little you can do about database performance, except for when something obvious shows up on the report.
Best regards,
Nikolay

New HP EVA6000 SAN and now bad database performance problems

Hello,
we changed our SAN Hardware to a HP EVA6000 and moved all Data to there.
It is intended that storagesystem is for all Servers (File/Print, Exchange, Oracle Databases and MSSQL Databases).
According to the best practice papers of HP we created one big discgroup (fata harddiscs) and created virtual discs for our servers.
After doing this the database performance was terrible bad!
Especially the multiple random IO is absolutely worse.
As a first countermeasure we created a second discgroup on faster harddiscs and moved the database contents to there. We analyzed the IO and moved several database file to different virtual discs.
The performance is better now, but still not like the 4 year old SAN System!
Of course we questioned HP and even let them do a performance analysis but up to now we have no solution... The performance analysis report will be available on thursday.
Does anybody had the same experience or how did you configure the Database and EVA SAN to work with an appropiate performance?

Hi,
I'm not an Oracle person, but do work with EVA SAN's.
Your 48 FATA disk, disk group is capable of 4800 I/O operations per second [48 x 100] but the disks are only rated for 30% duty cycle and spin at 7200 rpm.
The 16 FC drive Disk Group I/O operations rating depends on the speed the disks spin at. 10K rpm disks are rated at 130 I/O per sec [2080 for the group] and if they are 15K rpm disks then 170 I/O per sec [2720 for the group]. Both are rated for 100% duty cycle.
I seem to recall having have read somewhere that Oracle prefers to have it's logs on separate storage to it's data.
If your shelves had the spare disk slots I would put in 72 GB 15K rpm disks up to the capacity required {+ overhead } + head room for reasonably predictable growth over the anticpated life of the equipment.
Here is a link to the HP Best Practice guide for EVA 4/6/8000's
http://h71028.www7.hp.com/ERC/downloads/4AA0-2787ENW.pdf
I hope this helps you understand the storage that you are working with a bit better. The old saying of "more heads make for better performance" is still true, however budget can have some affect in performance. ;-)
Jim

High performance in memory database supporting C++ and JAVA

Hi,
I am considering to build a Java compatible In memory database to speed up the application of JAVA. We have now the C++ version of the database, named MTDB, which is hundreds of times faster than phisical database. The in memory database runs on solaris, linux and windows system. MTDB supports load sharing, data distribution, and realtime data synchronizing. MTDB shows large capacity, great performance, stability, and scalability in telecom applications. Add JAVA access interface to MTDB is very valuable. Please send me email if you have interest in this.
Thanks.

Are you selling it? In which case advertisements are not appropriate here.
There are a bunch of existing in-memory databases for Java which offer significant performance enhancements - unless yours outdoes them considerably I think you'll find a very limited market.
Check out hsqldb.org for one example.

Administer datafiles, tablespaces and free space

Hi all,
Firstly i want to apologize if my English is not very good, and secondly for if there is such topic already. I tried the "search" but didn't find something that suits my requirements...
We are using Oracle9i.
Here is the (potential) problem:
I noticed that the some of the filesystems soon will run out of space.
Example:
/dev/vg07/lv_data7 71671808 68657701 2825782 96% /var/application/oracle/data07
/dev/vg07/lv_data6 71671808 58102994 12720766 82% /var/application/oracle/data06
/dev/vg07/lv_data5 71671808 69448894 2084034 97% /var/application/oracle/data05
/dev/vg07/lv_data4 71671808 67765635 3662100 95% /var/application/oracle/data04
/dev/vg06/lv_data3 71675904 65840498 5470722 92% /var/application/oracle/data03
/dev/vg05/lv_data2 71675904 65730929 5573472 92% /var/application/oracle/data02
/dev/vg03/lv_data1 70656000 65924095 4436216 94% /var/application/oracle/data01
/dev/vg10/lv_data10 292929536 7301210 267776605 3% /var/application/oracle/data10
/dev/vg10/lv_data11 292929536 0 292929536 0% /var/application/oracle/data11
Lets for example take a look at data05:
/var/application/oracle/data05 > ls -l
There are many tablespaces and some of their datafiles there.
-rw-r----- 1 oracle dba 367005696 Jul 21 13:19 e1_data_d_d_11.dbf
-rw-r----- 1 oracle dba 209719296 Jul 21 13:20 e1_data_d_d_12.dbf
-rw-r----- 1 oracle dba 1022365696 Jul 21 13:19 m_data_d_07.dbf
-rw-r----- 1 oracle dba 812650496 Jul 21 13:19 m_data_d_08.dbf
-rw-r----- 1 oracle dba 707792896 Jul 21 13:19 m_data_d_09.dbf
-rw-r----- 1 oracle dba 2097156096 Jul 21 13:19 e_data_d_30.dbf
-rw-r----- 1 oracle dba 734007296 Jul 21 13:20 e_data_d_38.dbf
-rw-r----- 1 oracle dba 629149696 Jul 21 13:20 e_data_d_39.dbf
-rw-r----- 1 oracle dba 419434496 Jul 21 13:20 e_data_d_40.dbf
-rw-r----- 1 oracle dba 812650496 Jul 21 13:25 m1_data_d_03.dbf
-rw-r----- 1 oracle dba 393220096 Jul 21 13:27 m1_data_d_04.dbf
-rw-r----- 1 oracle dba 209719296 Jul 21 13:20 m1_data_d_05.dbf
-rw-r----- 1 oracle dba 2097160192 Jul 21 13:19 m1_data__03.dbf
-rw-r----- 1 oracle dba 2097160192 Jul 21 13:39 m1_data__04.dbf
-rw-r----- 1 oracle dba 2097160192 Jul 21 13:19 m1_data__05.dbf
-rw-r----- 1 oracle dba 2097160192 Jul 21 13:39 m1_data__06.dbf
-rw-r----- 1 oracle dba 2097160192 Jul 21 13:19 m3_data_d_d_03.dbf
-rw-r----- 1 oracle dba 2097160192 Jul 21 13:42 m3_data_d_d_04.dbf
-rw-r----- 1 oracle dba 1572872192 Jul 21 13:20 m3_data_d_d_05.dbf
drwxr-xr-x 2 root root 96 Nov 21 2006 lost+found
-rw-r----- 1 oracle dba 942678016 Jul 21 13:19 system02.dbf
-rw-r----- 1 oracle dba 72359936 Jul 21 13:20 system04.dbf
The situation is similar in the other filesystems - various tablespaces and a lot of datafiles.
Perhaps you've noticed that /data10 and /data11 are empty.
What is the best approach in this situation so that we start using those two filesystems .
Is moving datafiles from heavily used filesystems to the empty ones there is a good solution?
My idea is to make Oracle to automatically extend the tabespaces over the empty volumes and start creating the new datafiles there, but i am not sure how exactly to do that.
Any help will be appreciated.
Regards.
/R
Message was edited by:
user648100

Thanks for the quick reply.
MYH,
Yes, the tablespaces are autoextend on. Here is the information about one of them from the GridControl:
Name e1_data_d_d
Status ReadWrite
Type Permanent
Extent Management local
Storage
Allocation Type Automatic
Segment Space Management Automatic
Enable logging Yes
Block Size (B) 4096
I did not understand the second question - about partition a datafile?!? I can move datafiles but i'm still learning... never heard of partitioning a file, sorry.
Yas, i think the datafiles are Oracle Managed and oracle, after filling up a file, creates the new datafile automatically. I can create new file within the empty filesystems, but isn't it dangerous.
And with regards to same question - if i create a new datafiles, or move old datafiles - will Oracle start to use the new filesystems automatically?

DEFAULT TABLESPACE and TEMPORARY TABLESPACE for eBusiness Suite database

Hi:
What is the
DEFAULT TABLESPACE and TEMPORARY TABLESPACE
for eBusiness Suite database.
I want to create a new schema on the eBusiness Suite database.
Please help

DEFAULT TABLESPACE and TEMPORARY TABLESPACE
for eBusiness Suite database.
I want to create a new schema on the eBusiness Suite database.Create a new tablespace and temporary tablespace for your custom schema (you may use the existing temporary tablespace).
Integrating Custom Applications with Oracle Applications Release 11i [ID 176852.1]
Step By Step Guide to Creating a Custom Application in Applications 11i [ID 216589.1]
Thanks,
Hussein

Direct IO,Asynchronous IO and relationship with database performance

What is concept of Direct IO, Asynchronous IO from the DBA prospective? How does it relates to the database performance?
Any simple explanation will be highly appreciated. Thanks in advance.

918868 wrote:
What is concept of Direct IO, Asynchronous IO from the DBA prospective? How does it relates to the database performance?
Any simple explanation will be highly appreciated. Thanks in advance.yet another interview question from you?

Database Performance (Tablespaces and Datafiles)

Similar Messages

Maybe you are looking for