Database storage considerations having 10 hdds ?

Hi,
I must replace poor performing disks for my db2 sap database. Storage admin asigned 10 hdds from our corporate SAN (I mean I must decide in what RAID level disks will be configured(using SAN) ).
Config :
AIX 5.3
db2 8.1
SAP R/3
8 cpus
32 Gb RAM
I really must make the best out of these disks. So my thoughts are:
8 hdd's in RAID 5 for data
2 hdd's in RAID 1 for database logs.
How about my config ? Could I do better out of these 10 hdds ? Any suggestions ?
thank you in advance
Vilius

Hi Vilius,
What you described is reasonable. RAID-5 for sapdata's allows you to get a combination of performance (with striping to enhance the sequential prefetching) and data protection (with that extra disk).
And for log, RAID-1 mirror gives you immediate recovery (without time requires to rebuild from the bad disk like RAID-5) if one of disk goes bad. Are you sure though one disk size is big enough to hold all the active logs (including possible secondary logs)?
Another thing you may want to think about is once you run out of space for the current disk set, how do you add more. For example, if you add another 8 disks, to the sapdata, would you like to put them in the same stripe set (gives you 16 - 2 disk I/O throughput but requires time for table space rebalancing), or add them to the a new stripe set (which does not provide you the full 16 - 2 disks throughout, but avoid rebalancing).
Regards,

Similar Messages

DBCA can not use ASM for database storage due to following reason: Could no

Hi
I've a recently installed RAC11gR2(11.2.0.3) on HP-UX
I installed Grid Infrastructure with grid user:
Node1: # id grid
uid=1100(grid) gid=1000(oinstall) groups=1031(dba),1021(asmdba)
Node2: # id grid
uid=1100(grid) gid=1000(oinstall) groups=1031(dba),1021(asmdba)
I installed the Oracle Server with oracle user:
Node1: # id oracle
uid=1101(oracle) gid=1000(oinstall) groups=1031(dba),1021(asmdba),1033(oper)
Node2: # id oracle
uid=1101(oracle) gid=1000(oinstall) groups=1031(dba),1021(asmdba),1033(oper)
All related Grid Infrastructure is functioning well in both nodes:
$ crs_stat -t
Name Type Target State Host
ora.DATA.dg ora....up.type ONLINE ONLINE node03
ora.FRA.dg ora....up.type ONLINE ONLINE node03
ora....ER.lsnr ora....er.type ONLINE ONLINE node03
ora....N1.lsnr ora....er.type ONLINE ONLINE node04
ora.asm ora.asm.type ONLINE ONLINE node03
ora.cvu ora.cvu.type ONLINE ONLINE node04
ora.gsd ora.gsd.type OFFLINE OFFLINE
ora....network ora....rk.type ONLINE ONLINE node03
ora.oc4j ora.oc4j.type OFFLINE OFFLINE
ora.ons ora.ons.type ONLINE ONLINE node03
ora.scan1.vip ora....ip.type ONLINE ONLINE node04
ora....SM1.asm application ONLINE ONLINE node03
ora....03.lsnr application ONLINE ONLINE node03
ora....d03.gsd application OFFLINE OFFLINE
ora....d03.ons application ONLINE ONLINE node03
ora....d03.vip ora....t1.type ONLINE ONLINE node03
ora....SM2.asm application ONLINE ONLINE node04
ora....04.lsnr application ONLINE ONLINE node04
ora....d04.gsd application OFFLINE OFFLINE
ora....d04.ons application ONLINE ONLINE node04
ora....d04.vip ora....t1.type ONLINE ONLINE node04
$
The ASM instance are running well in both nodes
Connected with oracle user, the listener has registered perfectly the ASM instance in each node:
$
$ lsnrctl status
LSNRCTL for HPUX: Version 11.2.0.3.0 - Production on 16-MAY-2012 15:41:06
Copyright (c) 1991, 2011, Oracle. All rights reserved.
Connecting to (ADDRESS=(PROTOCOL=tcp)(HOST=)(PORT=1521))
STATUS of the LISTENER
Alias LISTENER
Version TNSLSNR for HPUX: Version 11.2.0.3.0 - Production
Start Date 16-MAY-2012 15:24:19
Uptime 0 days 0 hr. 16 min. 47 sec
Trace Level off
Security ON: Local OS Authentication
SNMP OFF
Listener Parameter File /u01/app/11.2.0/grid/network/admin/listener.ora
Listener Log File /u01/app/grid/diag/tnslsnr/node03/listener/alert/log .xml
Listening Endpoints Summary...
(DESCRIPTION=(ADDRESS=(PROTOCOL=ipc)(KEY=LISTENER)))
(DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=16.0.0.133)(PORT=1521)))
(DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=16.0.0.136)(PORT=1521)))
Services Summary...
Service "+ASM" has 1 instance(s).
Instance "+ASM1", status READY, has 1 handler(s) for this service...
The command completed successfully
$
$
DATA & FRA disk group are created:
$
$ echo $ORACLE_SID
+ASM1
$
$ asmcmd
ASMCMD>
ASMCMD> ls -lt
State Type Rebal Name
MOUNTED EXTERN N DATA/
MOUNTED EXTERN N FRA/
ASMCMD>
ASMCMD>
The big issue is... that the DBCA can not see the diskgroups. When I choose ASM Storage it can not see them. It shows the chart in blank and if i try to write the disk group, so the following message appears: can not use ASM for database storage due to following reason: Could not connect to ASM due to following error: ORA-01034 ORACLE not available
What can be passing in this case... ?
Thanks in advance

Thanks both
Really the problem initially was because of this:
[INS-06006] Passwordless SSH connectivity not set up between the following node(s): [node1, node2]. CVU may reports "PRVF-4007 : User equivalence check failed for user "user"" for same error
and I had to install node1 first and after node2 manually and perhaps for this the product was not well linked.
I resolved the problem establishing the keys with the wizard and after copying the following files from node1 to node2
/home/grid/.ssh -------> /home/grid/.ssh
authorized_keys
known_hosts
This is the procedure if you run the wizard at the node1.
The problem is that the wizard creates the keys incompletely...
Thanks

Database Storage location using DBCA

I'm trying to create a database. I would like to put the database on drive M and the logs on drive N.
I'm getting that if I want to do it during creation using DBCA I need to define the location on step 11 - Database Storage.
When I click on Datafile I then click on File Location Variables do I change ORACLE_BASE, ORACLE_HOME or both?
When I click on Redo Log Groups I then click on File Location Variables do I change ORACLE_BASE, ORACLE_HOME or both?
Do I do anything with the Control files location?

In that step you got File Name and File Directory menu.
Change the paths in File Directory menu.
Suppose you have create oradata folder in both M and N drives.
System01.dbf M:\oradata
redolog01a.log N:\oradata
suggestion:- Please multiplex controlfiles and redo log files.

Link to " Database Storage Consolidation with ASM White Paper " now fixed

I cannot access document "Database Storage Consolidation with ASM White Paper " on http://www.oracle.com/technetwork/database/index-100339.html.
The new link seems also to be broken http://www.oracle.com/technology/products/database/asm/pdf/db_storage_consolidation_wp%2012-05.pdf
Link is now working.
Edited by: P. Forstmann on 21 sept. 2010 08:51
Edited by: P. Forstmann on 21 sept. 2010 08:51

This link has been repaired.
Regards, OTN

Default Space for database storage

What the heck is that?
After I did a software update, this new feature showed up
under "Security".
1) It has a "Show Database" tab.
2) It has a "Drag down" for mb (from 1mb to 500mb)
bobo

Safari now comes with SQLLite database storage for offline applications. An example is here:
http://webkit.org/misc/DatabaseExample.html

Database storage

I have a confusion regarding the database storage mechanism.
I have created a tablespace
CREATE TABLESPACE TEST
DATAFILE '..\test01.DBF'
SIZE 152k
SELECT bytes FROM DBA_FREE_SPACE
WHERE tablespace_name ='TEST'
It shows 65536 bytes.
But it is smaller then 152k.
Why is it so??
Please help me out.

hello,
as you have not specified extent clause i think it is initial extent oracle creates of 64k (65536/1024=64k) as it is ...oracle system-managed extents
have a look at this
http://download.oracle.com/docs/cd/B19306_01/server.102/b14220/logical.htm#g23847
this is extract from oracle docs
For system-managed extents, Oracle determines the optimal
size of additional extents, with a minimum extent size of 64 KB.
If the tablespaces are created with 'segment space management auto',
and if the database block size is 16K or higher, then Oracle manages
segment size by creating extents with a minimum size of 1M.
This is the default for permanent tablespaces.experts please correct me.....if i am wrong
thanks and regards
VD
Edited by: Dixit on Sep 8, 2009 2:15 AM
Edited by: Dixit on Sep 8, 2009 2:16 AM
Edited by: Dixit on Sep 8, 2009 2:18 AM

Can I use Citadel as database in cRIO(having VXworks)

Hello,
Is it possible to have citadel as database in cRIO having Vxworks as OS.
I donot have any host pc in the developement setup.
I need to store data in cRIO by implementing any kind of data.
Kindly let me know what all are the drawback for citadel.

There is an SQLite library somewhere on one of the NI Developer Resources specifically ported to cRIO. That may be the solution for you.
Rolf Kalbermatter
CIT Engineering Netherlands
a division of Test & Measurement Solutions

Step for oracle database storage migration

Hi all
Need a help for database storage migration from old storage to new storage. Please advise and given step to migrate the database storage.
Database version: 9i,10g(ASM),10g and 11g (ASM)

Hi;
Your question is not clear. Migration from what to what?
For all migration question please review:
Master Note For Oracle Database Upgrades and Migrations [ID 1152016.1]
Different Upgrade Methods For Upgrading Your Database [ID 419550.1]
Migration of Oracle Database Instances Across OS Platforms[ID 733205.1]
How To Use RMAN CONVERT DATABASE on Source Host for Cross Platform Migration [ID 413586.1]
PS:Please dont forget to change thread status to answered if it possible when u belive your thread has been answered, it pretend to lose time of other forums user while they are searching open question which is not answered,thanks for understanding
Regard
Helios

Does Partitions take up database storage space

Thanks in advance.
I have tables with partitions that do not have data stored in them yet, does partitions without data stored in them take up database storage? How do you measure the storage used by a tables partitions that do not have data stored in them yet?

Is there a benefit to asking the same question over again? (If so, I'd love to hear it.)
Or would you be willing to continue with the original thread at Oracle Table and Index Partitions storage consumption?

UCM 11g Setting up Database storage of content

Hi,
We currently have a 7.5.2 instance running.
We are upgrading to a 11g instance, we are looking at storing the documents in the database, but we are having trouble setting this up.
Does anyone have the steps on how to does this? (finding oracles intructions hard to follow)
Also as a side note, I have seen post stating about having a consumption and contribution setup. Why is this?
Many thanks for all your help
James

Hi,
THis really is not so hard.
I strongly recommend you do this before you have any content in the system or it all gets confusing!
Login as weblogic
Go to Administration > Providers > DefaultFileStore > Info
Click Upgrade and then Update
Then restart UCM managed server
Login as weblogic
Go to Administration > Providers > DefaultFileStore > Info
Click on Edit
Click on Edit Rule
Check the JDBC storage radio
Click OK
you will get some confirmation stuff
restart UCM and you are done
Edited by: Tim Snell on 15-Dec-2010 09:41

Hardware (storage) consideration for RAC 10gR2?

Dear experts!
We have to implement availibility for our database at a customer - RAC 10gR2. We don't have really high database workload and large amount of data (max. 30 GB).
So I would prefer a 2 node RAC. What are the requirements for such a "low level" RAC - specially for storage? The two nodes are in different rooms so I need 2 "storage devices" to be safety against fire.
We have to use:
- OS: Win2003 or 2008
- Oracle Standard Edition --> ASM
So could you please give me praxis tips about hardware consideration (components needed)?
Thanks
Markus

jymarkusg wrote:
Do you mean if I have a two node cluster and one node fails my whole cluster is offline? In comparison to a three node cluster this works how? Could you please give me more detailed information about this?No, that's not what I meant. I was talking about potentially split brain problems. 2 nodes looses communication with one another. How does a node now know whether it is at fault, or the other node? Does it restart or not?
That said - one of the RACs we have, is a small 2 node RAC. And I bet there are many more like that out there. So it used. However... I would prefer to have at least 3 nodes.
One problem for example is failover load. Simplistic example. When 1 node fails, its processes fails over to the other node. Assuming an even spread of all processes, the other node now gets a 100% load increase. This can be nasty if the remaining node is not fully able to deal with that increase. In a 3 node setup, the remaining 2 nodes would have dealt with an easier managed 50% load increase.
I don't agree with you that linux is cheaper - it depends on the customer's infrastructure. It doesn't make sense to run a linux server in an only windows environment with only windows administrators! And oracle is available for many platforms, so this restriction shouldn't be a problem.People can learn. In fact, I would expect that administrators would want to expand their knowledge. That is IMO not the primary deciding factor. Linux is by far more a capable o/s in the cluster environment. One only have to look at the top 500 clusters in the world. 443 of these are running Linux. That is 88.6% of the biggest and baddest clusters in the world. How many are running Windows? A measly 5 clusters.
No comparison. No contest.
Dataguard was already in evolution - but therefor I need oracle enterprise edition. RAC in standard edition is for free (with the limitation of 4 nodes and 4 cpus - but this should meet our requirements :-)).Never done a SE RAC. Even our small 2 node baby RAC is a TB database. Creating a RAC for a 30GB database..? I'm not sure what the expectations are. This is really a minute database and can be full cached in server memory. Can easily be virtualised (and with that, fallen over between machines). So ito of performance, scalability and even redundancy, there are a number of other options besides RAC.
For interconnect I would use gigabit as minimum.I personally recommend using Infiniband. GigE is at its maximum. Ceiling has been reached. IB QDR is now available and provide 40Gbit pipes. When running IPoIB, you can have 65K MTU, aka super jumbo frames. Or, using Oracle 11g, you can use RDS instead which is a lot faster than IP.
As far as I see the biggest problem is the storage...Ito costs for a low cost solution, yep. Especially when only needing a few GB of storage as oppose to a few TB. And instead, the money spend on this, could perhaps be far better spend on a virtualised solution.

Database storage of multiple languages for name of a single entity

Surely I'm not the first person to have this problem, but I can't find many resources on how to solve this. Just some background on what I'm working with, I'm running a web application using the Struts framework with jdk 1.4 on a Weblogic 8.1 server connected to an Oracle 9i database, I'm using DAO for my persistence layer, and I'm not opposed to using any 3rd party software to assist in this task.
For example, say I have a database of DVDs, My application will retrieve the list of DVDs from the database to display to the user. Depending on the user's Locale I want to display the title of the DVD in that language.
This is a simple example, but the real-world application is much more complicated. The application is highly configurable meaning that through a configuration application, my internal users can add DVDs at any time, edit the title of the DVD, or add a new language that they want the title to display in, etc. When a customer logs in to the application front-end they will be able to change the language they want the application to display in at any time, and the application will need to reflect that change. The display of constant text like field labels and page text is easy using Struts but its this configurable dynamic text thats a bit more difficult.
Its obvious to me that I will need to store the multiple languages in the database in some way, I'm thinking of either storage in some xml format, or a new table hanging off the "dvd" table with languages for each dvd. Either of these options would work for my purposes, but I don't want to go reinvent the wheel when this same problem has arleady been solved. If someone could point me in the right direction I would appreciate it

One common way to handle this is to modify your key for the dvd title to include a locale identifier:
key dvdtitle
becomes
key locale dvdtitle
Now your db can hold multiple translations of the same dvd title.
Regards,
John O'Conner
Its obvious to me that I will need to store the
multiple languages in the database in some way, I'm
thinking of either storage in some xml format, or a
new table hanging off the "dvd" table with languages
for each dvd. Either of these options would work for
my purposes, but I don't want to go reinvent the
wheel when this same problem has arleady been solved.
If someone could point me in the right direction I
I would appreciate it

Move the database storage layout, please Help.

Hi Buddies,
My current database A (9i) 's storage layout is
File System SIZE
/yqa1/tech/oracle 8628 MB <-- Oracle software plus server's dump dirs only
/yqa7/appl/oracle 34521 MB <-data
/yqa8/appl/oracle 34521 MB <-data
we want to split existing 32G filesystem into separate chunks of 8G and map them into seperate filesystems
(for expamle: split /yqa7/appl/oracle into
/def1/appl/oracle [8G]
/ghi7/appl/oracle [8G]
/jkl23/appl/oracle [8G]
/xyz2/appl/oracle [8G]
for database A, finallly, we want to change the storage layout like this:
File System SIZE
/yqa1/tech/oracle 8628 MB <-- Oracle software plus server's dump dirs only
/yqa7/appl/oracle 8628 MB <-data
Is there any best practice to achive the purpose? or any frinds can provide practice documentation regarding to this?
Thanks a lot in advance.
Jerry

You'd need to take the tablespace offline
ALTER TABLESPACE tablespace_name OFFLINE;
before you move or copy the datafile to relocate it.
After completing the move or copy, you could use the
ALTER DATABASE RENAME FILE 'old_file_name' TO 'new_file_name' ;
for each of the datafiles in that tablespace.
Then, you can bring the tablespace online again
ALTER TABLESPACE tablespace_name ONLINE;

Database storage vs. File system

Hi,
We'll be setting up a new Oracle UCM 10g* environment in a production scenario - we have access to high-speed SAN for file storage, but we can also use an Oracle 11g database for file storage (using the File Store Provider) - the database's tablespace files will anyway be stored on the SAN. We're not quite sure which way to go - assuming we have complete ha/dr for both SAN and database solutions, which way is the best?
I'm also aware that if we don't select the webless option we'll anyway need a similar amount of storage on the file-system as the weblayout will contain a copy (or a web-rendition) of the checked in file - we could create a webless ucm, with inbound refinery only creating files in weblayout.
Any input on the matter would be appreciated.
*11g not chosen due to WebLogic licencing requirements for our client - anyone know if 11g will ever be released standalone?
EDIT - re: 11g licensing - sorry about that i misunderstood the licensing issue of our client.
Edited by: Etienne Azzopardi on 04-Feb-2011 13:44

Sorry, but I will start my reaction from the last point:
*11g not chosen due to WebLogic licencing requirements for our client - anyone know if 11g will ever be released standalone?What do you mean? UCM license contains a restricted license of Weblogic EE for the purpose of hosting UCM. In other words, WLS is there for free (only it cannot host anything else). 11g will never be standalone - it will only support other application servers, namely, Websphere and JBoss (it follows the logic for other non-Oracle middleware products).
File system vs. database - there's been a couple of posts on this topic.
In a nutshell:
- with regard to the speed: unless you have really a lot of items (millions a month at least, maybe even millions a day), file system will be slightly faster or equal
- in Oracle DB 11.2 or higher you can use SecureFiles, which can effectively work with unstructured content. It can squeeze the disk space needed for storage, speed up operations (checkins, but also e.g. backups)
- with DB you will effectively have one system to manage - well, UCM still requires file system (shared storage for clusters), because some operations like conversions cannot run on the content in the database
- DB gives you further options like partitioning, hierarchical storage management, advanced security, etc. Your SAN storage can have some of these too.

Sharepoint 2010 Database Storage double in size

Hi,
I've noticed something interesting about content databases on SharePoint 2010. I appears that size of the database is double the size of the content. Is this how SharePoint is designed?
For example, the web application content database "Current storage used: 2837 MB" in Central Admin, but if I check the actual content database size in DB Server it shows 4223 MB.

1st we need to identify that which thing is causing this issue:
open the SQl management Studio and connect to the SQL server from here:
right click on the Database > Reports > Standard Reports > Disk
Usage By Table
Check which table is chewing all the space(i.e alldoc, audit, alldocstream etc). if it is audit then you have to think becuase audit table never count towards quota.
2nd) also check if there is any white/free space in the content db.
3rd) make sure not too much data in the Site collection Admin recycle bin(2nd stage), this is not counting towards the quota.
4th) run the below script to force the recalculate the space of the site collection.
$URL = "https://<spsite>/<sitecollection>"
$Site = Get-SPSite -identity $URL
$Site.RecalculateStorageUsed()
$site.Usage.Storage
please share the result.
Please remember to mark your question as answered &Vote helpful,if this solves/helps your problem. ****************************************************************************************** Thanks -WS MCITP(SharePoint 2010, 2013) Blog: http://wscheema.com/blog

Database storage considerations having 10 hdds ?

Similar Messages

Maybe you are looking for