Results 1 to 6 of 6
  1. #1
    Join Date
    Jan 2006
    Posts
    3

    Unanswered: Can not initialize a full refresh (SQL replication) on the apply side

    Hello

    I tried to do an initial full refresh but I failed - nothing happens (no error at all. Status of run apply = 0). Apply works without errors. Details of my environment:

    I have the following replication scenario: DB2 UDB apply is running on a Solaris machine (DB2 Version 8, fixpack 9a), capture on a z/OS machine (host).

    On the z/OS machine (capture side), the column values of the column "DISABLE_REFRESH" for all source tables are set to 0 therefore allowing a full refresh.

    I successfully updated the ASN.IBMSNAP_SUBSET table on the apply side
    as followed (as suggested by the Replication Center while performing an
    automatic full refresh):

    CONNECT TO TESTDB;
    UPDATE ASN.IBMSNAP_SUBS_SET SET ACTIVATE=0 WHERE SET_NAME='S_TEST' AND
    APPLY_QUAL='A_TEST' AND WHOS_ON_FIRST='S';

    UPDATE ASN.IBMSNAP_SUBS_SET SET LASTSUCCESS=NULL, SYNCHPOINT=NULL,
    SYNCHTIME=NULL WHERE APPLY_QUAL='A_TEST' AND SET_NAME='S_TEST' AND WHOS_ON_FIRST='S';

    UPDATE ASN.IBMSNAP_SUBS_SET SET ACTIVATE=1 WHERE SET_NAME='S_TEST' AND
    APPLY_QUAL='A_TEST' AND WHOS_ON_FIRST='S';
    COMMIT ;

    Then - after checking that all above COLUMNS are set to the above
    values - I successfully started the apply but the full refresh didn't
    take place. Logfile:

    2006-01-24-16.15.30.394515 <CPCIMPC(08/07)> ASN1045I APPLY "A_TEST" :
    "Initial". The Apply program was started using database "TESTDB".

    2006-01-24-16.15.31.038834 <CPREST(01/00)> ASN1044I APPLY "A_TEST" :
    "WorkerThread". The Apply program will become inactive for "0" minutes
    and "50" seconds.
    ...

    2006-01-24-16.34.21.725226 <CPREST(01/00)> ASN1044I APPLY "A_TEST" :
    "WorkerThread". The Apply program will become inactive for "2" minutes
    and "0" seconds.

    There were no errors in the apply logfile and checking the
    ASN.IBMSNAP_APPLYTRAIL shows that no rows were inserted or deleted in
    the target tables (STATUS = 0, FULL_REFRESH = N):

    I checked the host side (where capture is running): the tables are not
    empty and my apply-process is activated.

    Does anybody have an idea why the initial full refresh doesn't take
    place? Are on the Capture side some other fields necessary exept DISABLE_REFRESH = 0 in order to process a full refresh?

    Thank you very much for your help

    With kind regards

    John

  2. #2
    Join Date
    Aug 2001
    Location
    UK
    Posts
    4,650
    Don't remember exactly ... ASN.IBMSNAP_PRUNCNTL & ASN.IBMSNAP_REGISTER tables are a few tables to consider .

    Cheers
    Sathyaram
    Visit the new-look IDUG Website , register to gain access to the excellent content.

  3. #3
    Join Date
    Nov 2004
    Posts
    374

    full refresh

    ibmsnap register keeps the parameter as indicated disable full refresh
    if set to 0 : full refresh can occur
    any test done with applytrace active : to see more detail
    I think you can still specify asnapply ..... trcflow > out.file
    also check capture trace table to see if any error over there
    but full refresh is done by apply
    have you tried to force manual/automated refresh through replication center and see the statements they generate to do this
    for initial full refresh no updates are needed to any table, as apply is started this will detect new subscription and trigger the full refresh..
    hope this will help
    Best Regards, Guy Przytula
    DB2/ORA/SQL Services
    DB2 DBA & Advanced DBA Certified
    DB2 Dprop Certified
    http://users.skynet.be/przytula/dbss.html

  4. #4
    Join Date
    Jan 2006
    Posts
    3
    Re: full refresh

    Hello Przytula

    Thank you very much for your answer.

    I already tried your suggestion with the trcflow option (still works). No errors or hints in the output.

    And yes, I tried the automated refresh through replication center, i.e. I updated the LASTSUCCESS, SYNCHPOINT and SYNCHTIME columns with NULL. Afterwards, apply performed with STATUS=0 but no full refresh was performed (in APPLYTRAIL table, the column was "full_refresh = N", telling me that no full_refresh was performed).

    I also activated the asntrc tool for the apply side but didn't find any hints. The problem is that the apply runs without generating an error but it doesn't perform a full refresh. Everything seems alright. And without an error it's quite difficult to find out why the full refresh doesn't take place ;-)

    Any further suggestions?

    Thank you

  5. #5
    Join Date
    Nov 2004
    Posts
    374

    full refresh

    out of ideas.
    I presume, that modifications do not come through, as full refresh has not been executed. Do you have a support contract - maybe a pmr could help.
    any other subscriptions ? if none, you could try to start the capture with cold, this would also force a full refresh of all sets.. have you checked also the captrace table for any messages..
    Best Regards, Guy Przytula
    DB2/ORA/SQL Services
    DB2 DBA & Advanced DBA Certified
    DB2 Dprop Certified
    http://users.skynet.be/przytula/dbss.html

  6. #6
    Join Date
    Jan 2006
    Posts
    3
    Thank you very much for your help. The errors were on the capture side: not all correct values in the capture control tables were inserted/updated (our customer is responsible for the capture side). Now the full refresh worked perfectly.

    With kind regards

    John

Posting Permissions

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