ASM or file system

We are in the process of transition of the Oracle 9.2 to Oracle 11 g R1. There is an exchange of views concerning the use of ASM (we use the file system currently in our environment 9i). What are the benefits by using their ASM on filesystem?, would we see any performance gains/etc.? any information would be greatly appreciated

DarkFiBrE72 wrote:
What are the benefits by using their ASM on file system?

Rather easy choice IMO. ASM without any doubts or concerns.

ASM works with Oracle. Is designed for Oracle. Fully supports Oracle. You can't say that any system file cooked.

Used for well over 3 years now and never had a single problem attributed to the ASM goes down. And we have a couple of databases exceeding a TB in size.

Would we see any performance gains/etc?

No! How the software layer of storage/volume manager can do the rotation of the faster disks and controller heads read and write data faster?

ASM is a s/w layer. no h/w. What it can do, is provide the level s/w RAID - although it is better to do it at the level to support him. The RAID is "+ better + ', but this does not imply that ASM now offers a sudden increase in performance that did not previously exist.

That can make the ASM is auto balance of hot spots on the disks. This does not make the I/O faster, it distributes it more evenly.

So no, don't expect ASM to wear a cowboy hats, spurs, set up a huge white Stallion and shoot silver bullets to kill the monster formidable performance.

There is an exception to this topic-ASM for Exadata. But it is a version special s/w and only available for clients running Exadata storage nodes. In this case ASM (among others) dynamically built of blocks of data and provide performance e/s in this way gains.

Bottom line is, however, that ASM is not something that makes the I/O to go faster. You want fast I/O? You need to watch all I/O subsystem (including the DSO is only a small part) - which is btw what is an Exadata storage node and why it offers incredible performance of I/o.

Tags: Database

Similar Questions

  • ASM cluster file system and df h

    Hello Oracle Community,

    I was wondering why my ASM cluster file system showing 96% (25 GB free) used, even if I moved and deleted (200 GB +) I still exports datapump see 96%.  When running the - h - max-depth = 1, it shows the correct value of 216 GB used.  But this isn't until hours later, when you run df h even once, I finally see the free space on the drive, which I was expecting.  The system is a RAC run red hat 6.5 with Oracle SE 11.2.0.4.

    [shared me@db02] $ - h - max-depth = 1
    from: cannot read directory '. / lost + found ': Permission denied
    64 k. / lost + found
    1.9 G. / datapump
    129. M / temp
    53. M / ctrlfilebkup
    214. G / dailyexport
    216G.
    [shared me@db02] $ df-h

    / dev/ASM/acfsvol-287 548 G 524 G 25 G 96% / u05/shared

    Could someone please clarify why this is happening.

    Thanks to you all!

    Found the answer:

    http://www.LinuxQuestions.org/questions/Linux-General-1/different-results-in-du-and-DF-841145/

  • Copy the files of the ASM disk (file system)

    Hi all

    I want to copy the files from the DSO to the filesystem (separate drive). is it possible... ? If yes how can I do this...

    can someone give me a helping hand for this?

    Thanks in advance,
    Max

    Max wrote:
    Hello

    I use version 11 GR 2.

    Please see,
    http://download.Oracle.com/docs/CD/E11882_01/server.112/e16102/asm_util003.htm#sthref951

    HTH
    Aman...

  • RMAN to the ASM and network file system

    Hi all

    I use Oracle RAC 11 g R2 + ASM.
    I am new to RMAN, but I am trying to build daily tasks to do below:

    1. backup of my database to the FRA located on ASM diskgroup + FRA.
    2. at the same time I want to save my database for Network File System (NFS) to be taken by system administrators to an archive location that I can't access. Since according to my knowledge can't copy files to ASM for outside... so for this I'm planing to use the command format of the NFS mount point.
    3. on daily basis in the NFS file will be taken by a work (that I am not responsible for) to the Archive, and then it will be purged from this location.
    4. I am planing to have my windows of 15 days retention strategy.
    5. on every day, I am planing to cross check my backup and then delete the expired so that taken from NFS in the journal of the archives is deleted.

    The RMAN I am planing to run it on a daily basis are:
    backup incremental level 0 cumulative device type disk tag 'FULL_DB_INCR_LV0' database; -- this will be to the default location (+FRA)
    backup incremental level 0 cumulative device type disk tag 'FULL_DB_INCR_LV0' database format 'NFS/oracle/backup%U';--another copy to be taken to the archive location
    crosscheck backupset;
    delete expired backupset; -- this will delete the backups taken to the Archive location
    Now my questions:

    1 what good solutions? take account of the fact that I need backup on the file system of the operating system so that the jobs of system administrators take the backup of their archives.
    2. If a backup is needed, I use CATALOGUE START WITH to recover and then recover.
    3. because my retention period is 15 days, and the backup in NFS files will be deleted on a daily basis, so this retention will apply for backups in ASM, is this a problem?

    Hello

    Easy way is to use the tag. Find the number of backupset is a difficult task.

    Create a template for your tag with a date.

    How to use the Substitution Variables of orders RMAN [427229.1 ID]

    How to use the tag name of RMAN with different attributes or variables. [580283.1 ID]

    You can check which backup doesn't have a copy on NFS using the query below:

    SELECT   BS.RECID BSKEY,
      BP.RECID KEY,
      BP.RECID RECID,
      BS.PIECES PIECECOUNT,
      BP.HANDLE FILENAME,
      BP.TAG TAG,
      BP.COPY# COPYNUMBER,
      BP.STATUS STATUS,
      BP.PIECE# PIECENUMBER
    FROM V$BACKUP_PIECE BP,
      V$BACKUP_SET BS
    WHERE (BP.SET_COUNT = BS.SET_COUNT
    AND BP.SET_STAMP    = BS.SET_STAMP)
    order by  bskey,copynumber
    
  • RAC ASM database and system files in the standby mode

    I have a database of ASM the knot two of the CARS and want to create a physical file system standby active database. This is supported by oracle? and you can share tips? Thank you

    I hope that's what you're looking for... See the below links

    http://pythianpang.WordPress.com/2009/07/07/Data-Guard-ASM-primary-to-filesystem-physical-standby-using-RMAN-duplicate/
    http://indiandba.blogspot.com/2011/03/create-physical-standby-on-normal-file.html

    Kind regards
    http://www.oracleracexpert.com
    RMAN Oracle 11 g new features
    http://www.oracleracexpert.com/2011/12/Oracle-recovery-managerrman-new.html
    Understand the Oracle RMAN Reporting
    http://www.oracleracexpert.com/2011/06/understand-Oracle-RMAN-reporting.html

  • Archived redo logs in ASM or OS file system?

    DB version: 10.2.0.4
    Version of the operating system: Sun Solaris SPARC 5.10


    We are on a node two CARS.
    Currently, we store logs archived in BONE cross mounted file systems. This is the recommended method?
    What we will have if we keep newspapers archived in ASM?

    Ideally, the location of the log archive must be on a shared volume. In solaris, you can use ASM as the shared diskgroup.
    When using the common diskgroup say + ARCH / (you can choose your own name diskgroup, but for the ease I posted like this) for archiving logs, two instances can write to the same location.
    Need a recovery and restore operation, you can simply restore the archivelogs ASM and run recovery.

    Well, there are some other well explained by the experts here. I just did a unique advantage.

    Kind regards
    Mahesh.

  • DRM data files is stored in the file system

    Hello team, in DRM I have 2 or 3 applications originally created 2 months ago. Now, I'm getting issue in DB space such as the data files are created the Local file system. Names of sample files are as below.


    / ddtrg7/Oracle/product/112/DBS/DRM_DATA

    /ddtrg7/oracle/product/112/dbs/DRM_INDEX

    /ddtrg7/oracle/product/112/dbs/DRM_PROPS

    /ddtrg7/oracle/product/112/dbs/DRM_TRANS

    /ddtrg7/oracle/product/112/dbs/DTRSPECDRM_DATA

    /ddtrg7/oracle/product/112/dbs/DTRSPECDRM_INDEX

    /ddtrg7/oracle/product/112/dbs/DTRSPECDRM_PR

    /ddtrg7/oracle/product/112/dbs/DTRSPECDRM_TRANS

    /ddtrg7/oracle/product/112/dbs/DTRDRM_DATA

    /ddtrg7/oracle/product/112/dbs/DTRDRM_INDEX

    /ddtrg7/oracle/product/112/dbs/DTRDRM_PROPS

    /ddtrg7/oracle/product/112/dbs/DTRDRM_TRANS


    We see in the newspapers that the tablespaces and data files have been created in ASM diskgroup even during the installation period. I think that since "/ dmtng7/oracle/product/112/dbs / ' is the place where DB settings file, it could be that the location has been provided as input during the installation/configuration. Ask for your help in this on how to get rid of this problem. We don't need this to be grown in the file system. Please indicate why these files are created.

    The kind assistance in the present.



    Thank you, roshi

    Problems you meet again.

    Thank you

    ~ KKT ~.

  • File system for FRA and external backups

    Oracle 11g R2.

    The FRA drive is formatted using Linux x 83 put on Diskgroup + FRA system.

    Backups are stored outside on an ext3 file system.

    Suppose now the database gets completely crashed and I need to recover by continuing external backups.  I will copy the external backups to the + FRA disk and then perform the recovery from there.

    My question is, given that external backups are on ext3 file system, while the FRA is the Oracle ASM filesystem, there will be a problem for the use of the file systems of the two difference?

    Thank you

    Scott

    scottjhn wrote:

    To summarize the issue:

    In my system, the database will be on ASM disk group + DATA (disk1 disk2).

    I will use another disk, disk3 (ext3 file system) for FRA, to organize data files backups and multiplexed logs (archive logs, online newspapers and control files).

    Earlier you said "I'll use the + FRA (ASM disk group) for the fast recovery area that holds that archiving multiplexed logs, recovery online logs, and control files.»  Nothing else. »

    Now, I need one more place to hold the copy of the multiplexed logs.

    Wouldn't not on the same + (ABOVE)?  Alternatively, use a disk full-fledged, disk4 (ext3 file system)?

    Thank you

    Scott

    First of all, the reason for multiplexing online redo logs and control files is not the same as that of multiplexing archivelogs. Do it online and the control file are actually a part of the database.  The database will not work without them.  This is why it is so important to have multiplex.  Archivelogs, on the other hand, are part of your backup strategy.  The database itself really doesn't care if you have any archivelogs at all.  The archivelogs multiplexing is much less common.  I have never multiplexed mine, but I do not have a task that runs every hour to copy any newly created on a server off-site.

    If your database is on ASM, at least a control file and a member of each group of redo log must be in ASM.  Technically, is not having to be, but it's really foolish not to do so.  Where to put the second copy?  Probably several schools of thought on this.  If your FRA is on ext3, so it seems a simple decision... a FRA copy, a copy in its default location in ASM.

  • Oracle file system

    Hello

    I want to know about file system in oracle 10g.
    Please advice me what is the best among OMF and manually.


    Kind regards
    Mark,

    OMF is not a file system but it's a way to manage the files of db oracle in a way more better. It is the basis of ASM in 10g which took a file system to manage not only the database files, but also files not also db 11g using the ACF. If you're on 9i, you are limited to the OMF format only, but if you are in 10g, ASM you should use to get a large number of benefits than to just the files is created and managed by oracle.

    HTH
    Aman...

  • Only biggest advantage of the DSO on Unix file system

    Version: 10 g 2, 11g, 11 2 GR

    If you were asked to mention the single biggest advantage of the DSO on Unix file system format to store the data files. What would you say?

    Hello

    ASM is a logical volume manager, it is not simply a set of file system.

    ASM allow you to connect or Disconnect (add or remove) disk when the Oracle database is running by using the simple SQL statement. I think it is a strong point for the DSO.

    More about ASM balance the load the e/s from disks to improve performance.

    This link will tell you much more:

    http://download.Oracle.com/docs/CD/E11882_01/server.112/e16102/asmcon.htm#OSTMG03601

    Hope this helps.
    Best regards
    Jean Valentine

  • Oracle EBS database to the file system to the DSO and cross platform copy.

    Hello


    I have an Oracle EBS 11.5.10.2 with Oracle 10.2.0.4 on Solaris database. I need to copy this based on RHEL 5.3 Linux with the same version but the file system is ASM. Please suggest what that all available options to move system files to ASM based DB. I just need to move on the DB and does not worry at this point the Application layer. Please suggest what that all available options to move system files to ASM based DB.


    Thank you
    -Samar-

    Samar,

    Please check your email, if you have got enough information please mark this thread as answered.

    Thank you.

  • Is OFCS (Oracle Cluster File System), a real file system?

    Hi all
    Is this system of files of cluster oracle something slightly FAT23 or NTFS, ext3 Linux for Windows?
    We install OS on FS, but with OCFS, we install the Oracle database on it instead of operating system?

    As usual
    ----
    Applications
    ----
    OPERATING SYSTEM
    ----
    File system
    ----
    Physical drive

    But the ocfs is something like:
    ----
    Oracle
    ----
    OCFS
    ----
    OPERATING SYSTEM
    ----
    File system of the operating system
    ----
    Physical drive

    Is this fair?

    OCFS is a clustered file system. You do not use it for operating system partitions (for example / opt) but only for the file systems that will be shared between servers in a RAC cluster to contain database files.

    You install RAC? If so, ASM is a better option than OCFS. If this isn't the case, then you have really not OCFS.

  • Confusion with the file system OCFS2 OCR and voting disk RHEL 5, Oracle11g,

    Hi all

    I am trying to install the Oracle 11 g 3 node RAC database

    The environment on which I must make this implementation is as follows:
    Oracle 11g.
    Red Hat Linux 5 x 86
    Oracle Clusterware
    ASM
    Storage EMC
    250 GB of disk storage.
    SAN

    From now on, I am installing Oracle Clusterware on 3 nodes.

    I performed these tasks for cluster installations.

    1 configure the kernel parameters
    2 configure the user limits
    3 edit the /etc/pam.d/login file
    4 set up the groups and users to the operating system for Oracle Clusterware
    5. configure the environment Clusterware Oracle owner
    6. install the rpm CVUQDISK package
    7 configure the Hosts file
    8. check the network configuration
    9 configure SSH on all the Cluster nodes (equivalence of the user)
    9 enable SSH on all the Cluster nodes (equivalence of the user)
    10 installing Oracle Cluster File System (OCFS2)
    11. check the Installation of the (OCFS2) Oracle Cluster file system
    12 set the OCFS2 (/ etc/ocfs2/cluster.conf)
    13 set up the stack of Cluster O2CB for OCFS2

    BUT, here again the after that I'm a little confused on how to proceed. The next step is to format the disk and mount the OCFS2, directories of software... so, create etc.
    I asked my administration system to provide me with two partitions so that I could get in shape with the OCFS2 file system.

    He wrote telling me.

    * "Is what you want before I do it?" *

    * / dev/emcpowera1 is 3 GB and OCFS2.* format
    * / dev/emcpowera2 is 3 GB and OCFS2.* format

    * Those who are big enough for you? If not, I can re-size and re - format the *.
    * before I climb on the servers.*
    * the SAN is shared storage. / dev/emcpowera is one of the three LUNS on *.
    * storage is shared, and it is 214 GB. At the moment there are only two *.
    * partitions on it - the ones I have listed below. Can I repartition the LUN everything *.
    * as you want it.*
    * Where do you want that these climbs to: *.

    * / dev/emcpowera1 *.
    * / dev/emcpowera2 *.

    * I thought if this mounting technique would work as follows: *.

    * emcpowera1: / u01/shared_config/OCR_config *.
    * emcpowera2: / u01/shared_config/voting_disk *.

    * Let me know how you would like that mounted. » *

    Please recommend me what I should pass so that I can ask him exactly the same.

    My second question is, as we use the DSO, for which I'm going to configure ASM after clusterware installation, do I install Openfiler?
    PLS, check the environmental information that I provided above and make recommendations.

    As of now, I am using Jeffery hunters guide to install the entire installation. You think the installation guide everything is fine with my environment?
    http://www.Oracle.com/technology/pub/articles/hunter_rac11gr1_iscsi.html?rssid=rss_otn_articles

    Kind regards
    MK

    Hello Lawrence,.

    Thank you.

    Mark this question as answered.

    See you soon,.

    Rodrigo Mufalani
    http://mufalani.blogspot.com

  • File system, verify or repair failed during the installation of Sierra

    I downloaded Sierra and when he tried to install I get an error "file system check or repair a failure", how can I resolve this problem and install Sierra?

    Start of recovery

    and open disk utility:

    -Look at the Smart function. If other verified tham means disc is a failure. The disk may be failing, even if the status is verified.

    -If she tells checked then repair/First Aid disk. If it can be repaired and then try reinstalling the Sierra.

    If it is repairable, replace the drive

  • Unable to write the email in the box to the letters. Make sure the file system allows write privileges, and you have enough disk space to copy the mailbox.

    How to make sure the file system gives me write rights? This started last night on my Dell desktop running Windows 10, I also a Dell laptop running Windows 10 and my emails come on it without any problem, but the Office of my home computer is so I need to get my emails on this work. Please answer as soon as POSSIBLE with everything you might need me

    you have an anti virus? try to turn it off.

Maybe you are looking for

  • Why used my chocies bookmark tool bar stay in place. They keep deleting.

    Something happened to my Firefox. All this makes office crazy last week. When I bookmark a site now on my toolbar (which is intact for months) the next time I have open the browser, they are gone. And things seem different. I uninstalled the program

  • Satellite L650-11F - question about brightness of the display by using Win XP

    I apologize;I want to know is why in windows xp before you plug the adapter into the laptop (Toshiba L650-11F), screen light is bright, but when I drag the adapter, the display light is dark? And the driver is installed Please help me

  • 15 - ax004na: brief ripple effect on the screen - Misty

    Hello Initially, noticed that when returning in the Office of my browser, it is a ripple effect very briefly on the screen. It looks like an interference more than anything else that's not office real ripple, its more about the top of the image; fogg

  • Windows Defender or not?

    Under control panel / window Defender (WD) the message is that it is OFF and click to turn it on. A click does nothing and finally it times out. Here is the interesting part, I have an icon in the taskbar for Windows Defender and it WE. Click on this

  • WebVPN client SFR module removes the http packets

    Hi, I have configured the WEBVPN access to ASA 5512 with SFR module a long time ago and internal http links have been working great. After the ASA upgrade to 9.5 (2), module of firepower to 6.0.0 - 1005 and DefenseCentar to 6.0.0 (build 1005), I am u