Get Data Moving 1
Results 1 to 8 of 8

Thread: Oracle to SQL Migration - ORA-27091: unable to queue I/O ORA-17510

  1. #1
    Mick is offline Junior Member
    Join Date
    Nov 2017
    Posts
    2
    Rep Power
    0

    Oracle to SQL Migration - ORA-27091: unable to queue I/O ORA-17510

    Hi,


    Would anyone have any advice on why I am receiving the Oracle error below when attempting to migrate data from an Oracle 12c database (Oracle redo logs are on ASM - Solaris O/S) to SQL Server 2016.


    The Oracle Endpoint has all the ASM details, is using the Binary Reader method, has all the required privileges and tests sucessfully. I have tried multiple times with copying the redo logs to an alternate folder and not copying them, all tests fail with the same error. The Oracle DBA's have checked ASM and there are no issues, or database integrity issues.


    I am using Replicate Express.


    00006164: [SOURCE_CAPTURE ]E: OCI error 'ORA-27091: unable to queue I/O
    ORA-17510: An attempt to do I/O of size 512 to block 0 is beyond file size 131072. Logical block size: 512.
    ORA-06512: at "SYS.X$DBMS_DISKGROUP", line 422
    ORA-06512: at line 1' [122307] OCI error. (oradcdc_io.c:673)
    00006164: [SOURCE_CAPTURE ]E: Failed to read from ASM file with thread id '1' from block number '0', size '512' [20014] Internal error (oradcdc_io.c:675)
    00006164: [SOURCE_CAPTURE ]E: Failed to read from Redo log +ORADATA/TSTG/tstg_redo01c.log [20014] Internal error (oradcdc_redo.c:808)
    00006164: [SOURCE_CAPTURE ]D: Close Redo log '+ORADATA/TSTG/tstg_redo01c.log' (oradcdc_redo.c:222)


    Thanks for your advice.


    Cheers
    Mick

  2. #2
    stevenguyen is offline Senior Member
    Join Date
    May 2014
    Posts
    290
    Rep Power
    5
    look like a redo log issue,

    could you verify with your DBA that this file exist?

    +ORADATA/TSTG/tstg_redo01c.log

    Also, what version of Replicate are your running? 5.???



  3. #3
    Mick is offline Junior Member
    Join Date
    Nov 2017
    Posts
    2
    Rep Power
    0
    Hi Steven,

    Thanks, yes the redo logs exist in that location and the version I am using is 5.0.2.49, this was the trial version I downloaded about 2 weeks ago.

    Cheers
    Mick

  4. #4
    ISS DBA is offline Junior Member
    Join Date
    Apr 2018
    Posts
    3
    Rep Power
    0
    Was there ever any further information on this error message? I recently upgraded a database to 12c and when initiating BFile replication, I am getting the exact same errors in the Attunity replicate log.

    On my ASM instance, I find the following in the ASM alert log:
    Errors in file /u01/app/grid/diag/asm/+asm/+ASM/trace/+ASM_ora_40073.trc:
    ORA-17510: An attempt to do I/O of size 512 to block 0 is beyond file size 819200. Logical block size: 512.

    The contents of the trace file referenced are:
    Trace file /u01/app/grid/diag/asm/+asm/+ASM/trace/+ASM_ora_40073.trc
    Oracle Database 12c Enterprise Edition Release 12.2.0.1.0 - 64bit Production
    Build label: RDBMS_12.2.0.1.0_LINUX.X64_170125
    ORACLE_HOME: /u01/app/grid/product/12.2.0/grid
    System name: Linux
    Node name: ISSUSHDB01.region.iss.biz
    Release: 3.10.0-514.28.1.el7.x86_64
    Version: #1 SMP Fri Jul 7 12:58:19 EDT 2017
    Machine: x86_64
    Instance name: +ASM
    Redo thread mounted by this instance: 0 <none>
    Oracle process number: 30
    Unix process pid: 40073, image: oracle@ISSUSHDB01.region.iss.biz

    *** 2018-04-02T13:33:35.542915-05:00
    *** SESSION ID:(170.25646) 2018-04-02T13:33:35.542940-05:00
    *** CLIENT ID:() 2018-04-02T13:33:35.542946-05:00
    *** SERVICE NAME:() 2018-04-02T13:33:35.542950-05:00
    *** MODULE NAME:(repctl@yb-iss-3 (TNS V1-V3)) 2018-04-02T13:33:35.542955-05:00
    *** ACTION NAME:() 2018-04-02T13:33:35.542960-05:00
    *** CLIENT DRIVER:() 2018-04-02T13:33:35.542965-05:00

    2018-04-02 13:33:35.542 : [ CSSCLNT][ CSSCLNT]clsssinit: initialized context: (0x7f527daf90f0) flags 0x115
    2018-04-02 13:33:35.543 : [ CSSCLNT][ CSSCLNT]clsssterm: terminating context (0x7f527daf90f0)
    2018-04-02 13:33:35.603 : [ CSSCLNT][ CSSCLNT]clsssinit: initialized context: (0x7f527daf90f0) flags 0x115
    2018-04-02 13:33:35.604 : [ CSSCLNT][ CSSCLNT]clsssterm: terminating context (0x7f527daf90f0)
    *** CLIENT DRIVER:() 2018-04-02T13:33:35.627664-05:00

    2018-04-02 13:33:35.627 : User SYS not found in OCR.

    *** 2018-04-02T13:33:46.864886-05:00
    ORA-17510: An attempt to do I/O of size 512 to block 0 is beyond file size 819200. Logical block size: 512.



    Not sure the nature of the "User SYS not found in OCR", but I do not find any issues with ASM and this database is working fine. It is a standalone Oracle restart database, not RAC.

    Any ideas?

    Thanks,
    Chris

  5. #5
    Hein is offline Senior Member
    Join Date
    Dec 2007
    Location
    Nashua, NH - USA.
    Posts
    135
    Rep Power
    11
    Quote Originally Posted by ISS DBA View Post
    Was there ever any further information on this error message? I recently upgraded a database to 12c and when initiating BFile replication, I am getting the exact same errors in the Attunity replicate log.

    On my ASM instance, I find the following in the ASM alert log:
    Errors in file /u01/app/grid/diag/asm/+asm/+ASM/trace/+ASM_ora_40073.trc:
    ORA-17510: An attempt to do I/O of size 512 to block 0 is beyond file size 819200. Logical block size: 512.
    Mick, did your issue ever come to full resolution? What where the details?

    Dear "ISS DBA",
    Just to be clear, you are not working with "Mick", merely reporting a same error message. Correct?

    You provided some excellent detail, but there some information missing, for starters the exact Replicate Version.
    And the exact Oracle Source reader configuration would likely be needed also.
    Best to EXPORT the TASK, to get that data and/or submit the reptask_<taskname>__YYMMDDHHMMSS.log

    Do you have a support contract?
    If so, then it would be best to open a support case with that data.

    I have a vague recollection of Block '0' not existing for ASM reads, but being valid for a file copied away from ASM.
    Maybe there is a source endpoint configuration issue creating an incorrect combination of settings. Please review and provide if need be.

    Also, there have been some fixed with the Oracle Bfile reader in recent Replicate updates.
    Therefor the exact version being used is (even more) important, and retesting under the most recent version is likely required.

    Good luck!
    Hein

  6. #6
    ISS DBA is offline Junior Member
    Join Date
    Apr 2018
    Posts
    3
    Rep Power
    0
    Hi Hein, thank you for the response.

    This is a separate issue from Mick's. I just started experiencing this recently with a 12c database upgrade and found this post. Was hoping for a quick-hit resolution.

    I am using Attunity Replicate 4.0.9.132. We received it through an application bundle so do not have direct access to Attunity support, although I have already requested our provider to open a case for this. I also have one with Oracle just to see if there are things I can check/verify on the Oracle side.

    I am uploading my exported task just in case there is something amiss. I am unable to determine anything.

    Thanks,
    Chris
    Attached Files Attached Files

  7. #7
    Hein is offline Senior Member
    Join Date
    Dec 2007
    Location
    Nashua, NH - USA.
    Posts
    135
    Rep Power
    11
    ISS DBA,
    Thanks for the JSON. It shows a very basic setup.
    You may want to try with: - Replicate has file-level access to the redo log files

    I believe that Replicate 4.0 did not have good ASM documentation
    Here is a (temporary) link to Appendix K from a recent Replicate userguide: "best Practices when Working with Oracle ASM" :

    Now did this ever work? With older Oracle? With Logminer?

    We should really first try this under a recent (5.5, 6.0, or even 6.1) Replicate version.
    I talked this over with the support team, and unfortunately cannot share a link here.

    Since you appear to be working in test, maybe you can just ask the vendor to ask Attunity for a link to a recent Replicate distribution kit under their support agreement.

    Good luck!
    Hein

  8. #8
    ISS DBA is offline Junior Member
    Join Date
    Apr 2018
    Posts
    3
    Rep Power
    0
    Thank you for all your responses Hein! They are very informative and helpful.

    I do have ASM redo using BFILE working fine on 11g. As you point out, it is probably compatibility between my version of replicate and 12c.

    LogMiner is working fine with 4x and 12c, but I can't stand the constant LogMiner session messages in the Oracle Alert log. Plus, performance is better with BFile, so I want to avoid LogMiner.

    We are planning a replicate upgrade to 5x soon. In the meantime, I may test with the suggestions from the best practices reference as it I need to get the file-level access worked out anyway.

    Thanks again!

    Chris

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •