Error source endpoint

[suvbin]

Hi team,

Getting below issue at source side -- oracle. 

OCI error 'ORA-15173: entry 'CECXA23' does not exist in directory '/'

ORA-06512: at "SYS.X$DBMS_DISKGROUP", line 518

Could you please let me know what is this error all about and how can it be resolved. 

Thanks.

[john_wang]

Hello @suvbin ,

Thanks for reaching out to Qlik Community!

This is a pure Oracle Logminer issue. We'd like to understand further the reason Oracle LogMiner is chosen rather than Replicate Log Reader. And:

1- please make sure the account have enough privileges, check out User Guide Required permissions .

2- It's better you work with your Oracle DBA to troubleshoot the issue. For example use a privileged account to run a pure LogMiner script to confirm if Logminer works in the env or not.

3- What's the Oracle version please?

Hope this helps.

Regards,

John.

[suvbin]

Hi John,

Thank you for the update. Replicate Log reader is been used. and oracle client version is 19.3.0.0.0.

[john_wang · Accepted solution]

Hello @suvbin ,

Thanks for the feedback. "DBMS_DISKGROUP" is an Oracle 'internal package' which is used to access ASM redo log files. Please take care from Oracle 11g Release 2 (11.2.0.2), Replicate must be granted the SYSASM privilege in order to access the ASM account. 

1- Please verify if any ASM Trace file generated during the time to get some clue

2- Use the same account to verify if you get same error in SQLPlus. A sample:

sqlplus asmuser/mypwd@\"192.168.33.179:1521/+ASM\" as sysasm set serveroutput on; declare v_handle number; v_pblksz number; v_fsz number; begin dbms_diskgroup.open('+DATA/ORCL/03A8996EAFFC5089E063B321A8C036F1/DATAFILE/system.272.1145725829','r',2,8192,v_handle,v_pblksz,v_fsz); dbms_output.put_line('open file sucess file,file handle NO. '||v_handle); dbms_output.put_line('open file sucess file,file file size is'||v_fsz); dbms_diskgroup.close(v_handle); end; /

     Where "asmuser/mypwd" is account name and its password. "192..." is the Oracle Server IP Address. "+DATA/ORCL/03A8996EAFFC5089E063B321A8C036F1/DATAFILE/system.272.1145725829" is one redo log file. The above script can be used to simulate the redo log file access via ASM.

Hope this helps.

Regards,

John.

[SushilKumar]

Hello team,

If our response has been helpful, please consider clicking "Accept as Solution". This will assist other users in easily finding the answer.

Regards,

Sushil Kumar

[suvbin]

Sure Sushil. Will do it . 

[suvbin]

This error is still persist. Whenever reload is performed, it will run fine and after few minutes, it will throw the below error. As per DBA , the logs only exist for 6 hours and also entry 'CECXA23' already exist in the directory. But why again facing the same issue.

00012340: 2023-12-02T18:35:43:74923  [SOURCE_CAPTURE  ]T:  Executing open ASM file statement 'declare filetype varchar2(100);                    lblksize number; filesize number; handle number; pblksize number;                          begin dbms_diskgroup.getfileattr('+RECOC1/CECXA23/ARCHIVELOG/2023_12_02/thread_1_seq_22804.13716.1154515161', filetype, filesize, lblksize);                              dbms_diskgroup.open('+RECOC1/CECXA23/ARCHIVELOG/2023_12_02/thread_1_seq_22804.13716.1154515161', 'r', filetype, lblksize, :1, pblksize, filesize);                      :maxblknum := filesize; :lblksize := lblksize; end;'  (oradcdc_io.c:837)

00012340: 2023-12-02T18:35:43:74923  [SOURCE_CAPTURE  ]T:  OCI error 'ORA-15173: entry 'CECXA23' does not exist in directory '/'

ORA-06512: at "SYS.X$DBMS_DISKGROUP", line 518

ORA-06512: at line 1' [1022307]  (oradcdc_io.c:853)

00012340: 2023-12-02T18:35:43:74923  [SOURCE_CAPTURE  ]T:  Failed opening ASM file '+RECOC1/CECXA23/ARCHIVELOG/2023_12_02/thread_1_seq_22804.13716.1154515161', thread id '1' [20014]  (oradcdc_io.c:856)

00012340: 2023-12-02T18:35:43:74923  [SOURCE_CAPTURE  ]V:  OCI call returned OCI_NO_DATA (e:\jenkins_slave\workspace\Replicate-2022.11.0-Win-x64\replicate\repsrv\endpoints\oracle\BinaryReader\oradcdc_thread.c - 977)  (oradcdc_redoutil.c:300)

00012340: 2023-12-02T18:35:43:74923  [SOURCE_CAPTURE  ]V:  Going to retrieve online REDO log with sequence '22804', thread '1'  (oradcdc_thread.c:808)

[john_wang]

Hello @suvbin ,

Thanks for the feedback. Would you please run the script block (copying it from the task log file) with in SQLPlus (the sample was in my previous comment) to verify the entry 'CECXA23' is available or not?

Good luck with resolving the issue,

John.

[suvbin]

DBA replied with 

we are able to directory ASMCMD> pwd
+RECOC1/CECXA23/ARCHIVELOG
ASMCMD>

And also archivedlogs are available for 6 hours and archived later.

***************************

When i got the below error ..  they replied that logs are present...

00003672: 2023-12-01T03:40:12 [SOURCE_CAPTURE  ]E:  OCI error 'ORA-15173: entry 'CECXA23' does not exist in directory '/'

ORA-06512: at "SYS.X$DBMS_DISKGROUP", line 518

ORA-06512: at line 1' [1022307]  (oradcdc_io.c:853)

00003672: 2023-12-01T03:40:12 [SOURCE_CAPTURE  ]E:  Failed opening ASM file '+RECOC1/CECXA23/ARCHIVELOG/2023_11_30/thread_1_seq_22454.4057.1154293879', thread id '1' [20014]  (oradcdc_io.c:856)

00003672: 2023-12-01T03:40:13 [SOURCE_CAPTURE  ]I:  Failed to open Redo log '+RECOC1/CECXA23/ARCHIVELOG/2023_11_30/thread_1_seq_22454.4057.1154293879' (sequence 22454, thread 1), retry till Redo log with this sequence could be opened successfully  (oradcdc_redo.c:737)

Reply from DBA

ARCHIVELOG  MIRROR  COARSE   NOV 30 21:00:00  Y    thread_1_seq_22454.4057.1154293879

ARCHIVELOG  MIRROR  COARSE   NOV 30 21:00:00  Y    thread_1_seq_22455.1342.1154295579

ARCHIVELOG  MIRROR  COARSE   NOV 30 21:00:00  Y    thread_2_seq_27829.3502.1154295547

ASMCMD>

[john_wang]

Hello @suvbin ,

Thanks for your update. It's hard to tell from the piece of information. It's better you can run the same block of scripts( in my previous comments) during the error time to see if you get same error.

If you still cannot manage it solved, please:

1- check with DBA if there is any clue in Oracle alert log etc as it's a pure Oracle error.

2- Open a support ticket and provide task Diag Packages with source_capture set to Verbose, support team would like to help you further.

Regards,

John.