This discussion is archived
6 Replies Latest reply: Jun 6, 2013 3:43 AM by Santosh Pradhan RSS

RMAN_ASM_STRANGE issue seen

Santosh Pradhan Explorer
Currently Being Moderated
Guys ,,
i got very strange issue at site :

oracle 11r2 enterprise edition

Primary database(cbsprod)
database size = 1.5 terabytes
9 diskgroups are there and each diskgroups containing only one disk of 200gb

We are creating one physical standby(cbsdr) .

CBSDR storage details :
same as primary database i.e
9 diskgroups are there and each diskgroups containing only one disk of 200gb
all the disks are free and usable , verified from asmcmd lsdg

But while restoring backup on cbsdr it is showing
ora 19504 = failed to create file 'DATA1'
ora 17502 = ora-17502 ksfdcre failed to create file
ora 15041 = diskgroup +data1 space  exhausted

,
after throwing above errors i checked that only 5 datafiles are restored on +DATA1 containing space arround
50 gb

space on cbsprod in +data1 is 150 gb utilized and 50 gb is free


m not understanding if there is 200gb free space available then why this error is comming on cbsdr if all the structure is same ..


trying to resove since 3 days and also got help from experts but not done yet
  • 1. Re: RMAN_ASM_STRANGE issue seen
    Hemant K Chitale Oracle ACE
    Currently Being Moderated
    How many channels are you using for the restore ?

    Is DG +DATA1 the only one that has datafiles restored ?  No datafiles restored to any other DG ?
    after throwing above errors i checked that only 5 datafiles are restored on +DATA1 containing space arround 50 gb
    Does "containing space" mean "used space for the 5 datafiles" ? What are the sizes of the datafiles that need to be restoerd to +DATA1 ?

    Do you have DB_UNIQUE_NAME defined for CBSDR ? The ASM path would follow from the DB_UNIQUE_NAME so datafiles would be restored to +DATA1/CBSDR if DB_UNIQUE_NAME is "CBSDR".



    Hemant K Chitale
  • 2. Re: RMAN_ASM_STRANGE issue seen
    MahirM.Quluzade Guru
    Currently Being Moderated
    Hi,
    Santosh Pradhan wrote:
    Guys ,,
    i got very strange issue at site :

    oracle 11r2 enterprise edition

    Primary database(cbsprod)
    database size = 1.5 terabytes
    9 diskgroups are there and each diskgroups containing only one disk of 200gb

    We are creating one physical standby(cbsdr) .

    CBSDR storage details :
    same as primary database i.e
    9 diskgroups are there and each diskgroups containing only one disk of 200gb
    all the disks are free and usable , verified from asmcmd lsdg
    You must check DATA1 diskgroup size both side.
    But while restoring backup on cbsdr it is showing
    ora 19504 = failed to create file 'DATA1'
    ora 17502 = ora-17502 ksfdcre failed to create file
    ora 15041 = diskgroup +data1 space  exhausted
    Did you sure DATA1 diskgroup size is 1.8 TB?
    I think you must check ASM Diskgroups on standby side.
    >
    ,
    after throwing above errors i checked that only 5 datafiles are restored on +DATA1 containing space arround
    50 gb

    space on cbsprod in +data1 is 150 gb utilized and 50 gb is free
    It means DATA1 diskgroup contains only 1 disk. (150 + 50 = 200 GB)
    I think you must add other disk to DATA1 diskgroup, too.
    Because you need >1.5 TB space.
    m not understanding if there is 200gb free space available then why this error is comming on cbsdr if all the structure is same ..


    trying to resove since 3 days and also got help from experts but not done yet
    Did you using same DISK Group name both side?


    Regards
    Mahir M. Quluzade
    http://www.mahir-quluzade.com
  • 3. Re: RMAN_ASM_STRANGE issue seen
    Santosh Pradhan Explorer
    Currently Being Moderated
    Hi guys,,

    Issue resolved ,,

    Actually i have follwing paramete setting
    db_create filedest ='+data1'

    after googling i found that , this is the default value which is created at the time of database creation(dbca) .

    I just edited above parameter as follows :
    alter system set db_create filedest ='' scope=spfile

    now plz consider below example
    i want to restore datafile 1 that is system datafile on standby database ,
    on primary the name is "system.256.4545458" BUT when i restore the same
    b'cause of asm , after restoration the name automatically changed to "system.256.4545562"

    If i restore database like this then all the datafile names would be mismatch between primary and standby so how it start managed recovery process
  • 4. Re: RMAN_ASM_STRANGE issue seen
    MahirM.Quluzade Guru
    Currently Being Moderated
    Can you paste here on standby side?
     
     select name from v$datafile; 
    How to created standby database? Did you using duplicate command RMAN ?


    Mahir
  • 5. Re: RMAN_ASM_STRANGE issue seen
    Shivananda Rao Guru
    Currently Being Moderated
    Hello,

    You are using OMF (Oracle Managed Files). Oracle will take care about the naming of the files. You do not have to worry, just set the db_create_file_dest parameter to the diskgroup where you need to have the datafiles placed.
    b'cause of asm , after restoration the name automatically changed to "system.256.4545562"

    If i restore database like this then all the datafile names would be mismatch between primary and standby so how it start managed recovery process
    system.256.4545562 ---> The number is an unique number which Oracle sets for each datafile. You do not have to worry about it.
    Once the restoration is completed and you have recovered the standby database, start the MRP on the standby database using the below query.
    SQL>alter database recover managed standby database disconnect from session;
    Santosh Pradhan      
         Newbie
         
    Handle:      Santosh Pradhan
    Status Level:      Newbie (15)
    Registered:      Feb 4, 2010
    Total Posts:      240
    Total Questions:      94 (67 unresolved)
    Please consider closing your threads by providing appropriate points if you feel that they have been answered.

    Regards,
    Shivananda
  • 6. Re: RMAN_ASM_STRANGE issue seen
    Santosh Pradhan Explorer
    Currently Being Moderated
    Hi guys ,,


    Mahir , i didnt use duplicate command for standby creation as Shivananda Rao explained OMF had taken care of the naming convention in my case resolved that doubt .

    For standby creation i did :
    1.Install grid on standby server
    2.created ASM instance and added sufficient diskgroups and disks as same as production structure .
    3.started ASM instance on standby
    4.taken RMAN backup from production
    5.and for restoration i used following command :
    run
    {
    allocate channel c1 device type disk;
    allocate channel c2 device type disk;
    allocate channel c3 device type disk;
    restore database ;
    recover database ;
    release channel c1;
    release channel c2;
    release channel c3;
    }

    and its done .....

    thnk u very much Shivananda Rao for gud explainations..

Legend

  • Correct Answers - 10 points
  • Helpful Answers - 5 points