Results 1 to 7 of 7

Thread: CDC State Aborted and

  1. #1
    Jputter is offline Junior Member
    Join Date
    Nov 2012
    Posts
    2
    Rep Power
    0

    Oracle CDC State stuck on Aborted

    Hi,

    We have used the CDC service to replicate tables from Oracle to SQL. Over the past 3 months its worked great.
    Although recently we added about 12 new tables to the process it runs fine and did so for a week.

    This morning it showed a status Aborted on the CDC Designer for Oracle. Tried restarting the service also tried resetting but every time it immediately go's to the Aborted state. The trace log was active and when I clicked on collect diagnostics it didn’t show any thing on the log with a status aborted. In most cases it actually popped up with a message asking met to wait for the log , skip or cancel. So this means I cant even attach a log file to the thread as it doesn’t actually get to create one.

    Also to note here is that when I switch between tabs on the CDC designer it gives an error "Key could not be found in Dictionary", "
    at System.Collections.Generic.Dictionary`2.get_Item(T Key key)
    at Attunity.SqlServer.XdbCdcConfig.Core.Managers.CDCM anagerImp.GetCaptureInstances(CDCInstanceInfo cdcInfo)"

    What we have picked up from some sites are that the "key could not be found" error is due to structural changes that happened from oracle's side although how do you know what changes were made if they took place. like an index that gest dropped or recreated.

    Still new to this product so any help here would be appreciated.
    Last edited by Jputter; 12-03-2012 at 03:41 AM. Reason: Title change

  2. #2
    Gadi.Farhat is offline Development Manager
    Join Date
    Mar 2009
    Posts
    255
    Rep Power
    11
    Hi,

    First, let me start with some questions:
    1. Did you do any manual changes to the change tables or the "cdc.*" tables?
    2. Did you drop or rename any of the change tables manually i.e. without dropping the capture instance?
    3. Do you have tables in the target change database, with the same name but, different schema?



    Now, to get the trace information of the specific CDC instance, select from the cdc.xdbcdc_trace table in the target CDC database and attach the results.


    Regards,

    Gadi
    To Find Out more About Attunity Technology:
    Attunity
    or:
    Contact Us

  3. #3
    Jputter is offline Junior Member
    Join Date
    Nov 2012
    Posts
    2
    Rep Power
    0
    Hi Gadi
    Thanks for the feedback :),

    Your first question seemed to have been the problem. One of the developers added primary keys to the tables that did not exist.
    Recreated the DB and it works.

    Seem to be getting an issue wit the log files though. This has been happening for a while.
    I am attaching the log for reference.

    CDC go's into a logger, suspended state and looking at the log file created by collect diagnostics I can see the redo logs piling up one by one.
    The error I get is a "failed to add redo log" (refer to log file).
    This only seems to happen every now and then but I'm afraid that changes might not be picked up in this case , or that it doesn't come out of the suspended state.

    Any thoughts on this?
    Note that the service has run fine with no glitches for about 3 days and just suddenly again started this today. The Oracle DBA's say no changes was made.

    Thank you
    Attached Files Attached Files

  4. #4
    Gadi.Farhat is offline Development Manager
    Join Date
    Mar 2009
    Posts
    255
    Rep Power
    11
    Hi,

    If you are still getting the failed to add redo log error, can you please attach a recent log file created by collect diagnostics and the result of running the following queries on the source Oracle database:

    Select * from v$log;

    Select * from v$logfile;

    Select * from v$archived_log;


    Regards,

    Gadi
    To Find Out more About Attunity Technology:
    Attunity
    or:
    Contact Us

  5. #5
    Sql arm is offline Junior Member
    Join Date
    Sep 2013
    Posts
    2
    Rep Power
    0

    Aborted status please help

    I have installed “Change Data Capture for Oracle by Attunity” in our Development environment in order to develop Prove of Concept for our data warehouse project.

    Testing the functionality with one table against my development Oracle database works great after trying to do the same against our development database that has over 170 tables the service stopped with the following error message “ABORTED”.

    The only record that I can find it’s in the MSXDBC database in the xdbcdc_trace table “ORACDC205E:The Oracle CDC instance MDL_CR_CHANGES was aborted.” And this error message in trace log "RUNNING","IDLE","ORACDC517E:Oracle Call Intreface (OCI) method failed: ORA-01882: timezone region not found”

    The product that shipped with SQL Server 2012 Enterprise /Development version limited to the number of tables ?


    I have tried to collect diagnostics logs and troubleshoot the problem to find out the any indications in the trace logs without any luck.

    I also have posted the issue on the “Attunity” forums no replays so far.

    Please see attached screenshot of the error message.

    Could you please advise how to capture the reason and resolve this issue?

  6. #6
    dempsterj is offline Junior Member
    Join Date
    Oct 2013
    Posts
    7
    Rep Power
    0

    CDC State goes to Aborted

    I am in the process of setting up a new test environment to connect to an Oracle database. I believe most of the permissions on the Oracle side are set up correctly, but I think I'm missing something. When I start the CDC service, it will run for about 2 minutes, updating the "Current change number", but then it aborts. Below are the last several log entries from the trace table.

    I have deleted the CDC instance and recreated it a couple of times, in case there was an issue in the SQL database. Unfortunately, I'm not seeing any error messages, just trace messages. I'm guessing there's something happening on the Oracle side, but I don't have enough detail to know what.

    Is there a step I missed or a security setting?
    /*
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Reading next record
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Positioning to first change event in log miner session
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:EOF reached
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Retrieving current SCN
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Reading next record
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Adding redo logs to log miner session.
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Redo log: /oradata/sunprod/redo/redo_002.dbf, Sequence: 76443
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Thread(1) - Start log miner session
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Positioning to first change event in log miner session
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Retrieving current SCN
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Reading Complementary records for large log miner SQL_REDO/SQL_UNDO

    */

  7. #7
    dempsterj is offline Junior Member
    Join Date
    Oct 2013
    Posts
    7
    Rep Power
    0
    Quote Originally Posted by dempsterj View Post
    I am in the process of setting up a new test environment to connect to an Oracle database. I believe most of the permissions on the Oracle side are set up correctly, but I think I'm missing something. When I start the CDC service, it will run for about 2 minutes, updating the "Current change number", but then it aborts. Below are the last several log entries from the trace table.

    I have deleted the CDC instance and recreated it a couple of times, in case there was an issue in the SQL database. Unfortunately, I'm not seeing any error messages, just trace messages. I'm guessing there's something happening on the Oracle side, but I don't have enough detail to know what.

    Is there a step I missed or a security setting?
    /*
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Reading next record
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Positioning to first change event in log miner session
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:EOF reached
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Retrieving current SCN
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Reading next record
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Adding redo logs to log miner session.
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Redo log: /oradata/sunprod/redo/redo_002.dbf, Sequence: 76443
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Thread(1) - Start log miner session
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Positioning to first change event in log miner session
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Retrieving current SCN
    TRACE ETL-BI12-01-T RUNNING IDLE ORACDC000T:Reading Complementary records for large log miner SQL_REDO/SQL_UNDO

    */
    Please disgard my earlier post. We restarted the database, and it appears everything is working now.

Tags for this Thread

Posting Permissions

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