JR50865: CDC GETS 'DOES NOT CONTAIN VALID REDO LOG FILE INFORMATION' ERROR MESSAGE BECAUSE A POTENTIAL ORACLE BUG.

APAR status

• Closed as program error.

Error description

• In a busy oracle environment where archived logs can be created
  quickly, customer can run into situation that CDC would failed
  to find archived log on a given scn. From the diagnostic trace
  we'd given to customer, it is shown that when executing the
  first query to get the first_change# from a given scn, we get a
  row back, and then when executing the 2nd query to find out all
  the redo log names from the given first_change#, we got no row
  back. If the first query had returned something, then the second
  query must also return some rows, and yet that is not the case,
  so this clearly shows that Oracle is somehow not able to return
  rows we're looking for, and this is an Oracle bug (likely due
  to Oracle log retention policy kicks in to remove old archived
  logs that caused query to v$archvied_log stored in control file
  to return no rows). However we were unable to prove this since
  customer was not willing to run our test program, so instead of
  waiting for proving the problem is Oracle bug we can work-around
   the problem by re-executing the 2nd query up to 3 times, then
  that should be sufficient to get around this problem. But just
  to be clear this is an unexpected oracle behaviour we're trying
  to work-around, it's not a bug within CDC.
  Error message: e.g.
  87620    2014-06-28 11:45:26.823    SHAREDSCRAPE LOG
  READER:3{302}
  com.datamirror.ts.util.oracle.OracleRedoNativeApi
  logEventCallback()    Redo log file information not found for
  subscription SHAREDSCRAPE. The Oracle database does not contain
  valid redo log file information corresponding to the log
  position '12324813747543.0.0.0.0.0.0' for the current database
  incarnation. Error detected in file OracleRedoNativeApi.java at
  line 1178. This may be caused by any of the following: a) The
  SCN portion of the log position 12324813747543.0.0.0.0.0.0 is
  invalid. b) The SCN associated with log position
  12324813747543.0.0.0.0.0.0 does not correspond to a current
  on-line redo log file and ARCHIVELOG mode is not enabled. c) The
   redo log file(s) corresponding to log position
  12324813747543.0.0.0.0.0.0 does not currently exist, exists but
  is inaccessible or exists and is corrupt. d) The control file
  contains invalid information. e) The database was shut down
  abort. Re-start replication will resolve this issue. f) There
  was error archiving redo logs because of insufficient disk
  space. Try to resolve this issue by doing the following and
  re-start replication: a) Change the log position to specify a
  correct SCN value. b) Examine the views V$LOG and V$ARCHIVED_LOG
   to determine a valid SCN for which a corresponding online or
  archived redo log file exists. Change the log position to
  specify a valid SCN for the current database incarnation. c)
  Restore a valid copy of the required redo log file(s) from
  backup. d) Ensure the disk space for archived redo logs is not
  full.
  

Local fix

Problem summary

• IIDR 10.2 and 10.2.1 for Oracle Redo replication fails with
  "Redo log not found" type error when unable to find the log name
  based on an SCN.
  
  This issue affects users running IIDR 10.2 Interim Fix 17 for
  Oracle Redo (and lower) and IIDR 10.2.1 Interim Fix 5 for Oracle
  Redo (or lower).
  

Problem conclusion

• This issue is fixed by applying the following interim fixes
  depending on the product version:
  - IIDR 10.2 Interim Fix 18 for Oracle Redo, or
  - IIDR 10.2 Interim Fix 6 for Oracle Redo.
  

Temporary fix

• Restart replication.
  

Comments

APAR Information

• APAR is sysrouted FROM one or more of the following:

• APAR is sysrouted TO one or more of the following:

Fix information

• Fixed component name

  IS CDC ORACLE

• Fixed component ID

  5725E30OR

Applicable component levels

• RA20 PSY

     UP

• RA21 PSY

     UP