Moneycontrol Brokerage Recos

Thursday, August 15, 2019

Error ORA-235 occurred during an un-locked control file transaction




Today, I was doing a new build of physical standby database on one of the Exadata Machine and pending archive logs were shipping and applying over standby at rapid speed. I had OMF/FRA configured where RMAN archivelog deletion policy was to set to APPLIED ON STANDBY. So, constantly archives were getting shipped, applied and was being deleted along to make free room inside the FRA for the incoming archives from the primary database resulting massive access and update in controlfile header.


I noticed the below alert log entry:


Thu Aug 15 06:53:57 2019 
Archived Log entry 4053 added for thread 1 sequence 738452 rlc 948642470 ID 0x0 dest 2: 
Thu Aug 15 06:53:59 2019 
Media Recovery Log +RECO/STBYDB/ARCHIVELOG/2019_08_15/thread_1_seq_738409.108786.1016347787 
Thu Aug 15 06:54:00 2019 
Archived Log entry 4054 added for thread 1 sequence 738450 rlc 948642470 ID 0x0 dest 2: 
Thu Aug 15 06:54:00 2019 
Media Recovery Log +RECO/STBYDB/ARCHIVELOG/2019_08_15/thread_2_seq_730363.180664.1016345945 
RFS[6]: Selected log 19 for thread 2 sequence 732775 dbid 1234567890 branch 948642470 
Thu Aug 15 06:54:03 2019 
Media Recovery Log +RECO/STBYDB/ARCHIVELOG/2019_08_15/thread_1_seq_738410.175606.1016347787 
Thu Aug 15 06:54:06 2019 
Media Recovery Log +RECO/STBYDB/ARCHIVELOG/2019_08_15/thread_1_seq_738411.50462.1016347787 
Thu Aug 15 06:54:06 2019 
Archived Log entry 4055 added for thread 1 sequence 738449 rlc 948642470 ID 0x0 dest 2: 
Thu Aug 15 06:54:06 2019 
Archived Log entry 4056 added for thread 2 sequence 732774 ID 0x7b963ba0 dest 1: 
Thu Aug 15 06:54:06 2019 
Error ORA-235 occurred during an un-locked control file transaction.  This 
error can be ignored.  The control file transaction will be retried. 

RFS[2]: Selected log 14 for thread 1 sequence 740920 dbid 1234567890 branch 948642470 
Thu Aug 15 06:54:07 2019 
Archived Log entry 4057 added for thread 1 sequence 740919 ID 0x7b963ba0 dest 1: 
Thu Aug 15 06:54:09 2019 
Media Recovery Log +RECO/STBYDB/ARCHIVELOG/2019_08_15/thread_2_seq_730364.86799.1016345945 
Thu Aug 15 06:54:10 2019 
Deleted Oracle managed file +RECO/STBYDB/ARCHIVELOG/2019_08_15/thread_1_seq_738283.114273.1016344287
Thu Aug 15 06:54:10 2019 
Deleted Oracle managed file +RECO/STBYDB/ARCHIVELOG/2019_08_15/thread_1_seq_738279.82856.1016344281
Thu Aug 15 06:54:10 2019 
Deleted Oracle managed file +RECO/STBYDB/ARCHIVELOG/2019_08_15/thread_1_seq_738282.6538.1016344287
Thu Aug 15 06:54:10 2019 
Deleted Oracle managed file +RECO/STBYDB/ARCHIVELOG/2019_08_15/thread_1_seq_738278.165959.1016344279
Thu Aug 15 06:54:10 2019 
Deleted Oracle managed file +RECO/STBYDB/ARCHIVELOG/2019_08_15/thread_1_seq_738280.23165.1016344283
Thu Aug 15 06:54:10 2019 
Deleted Oracle managed file +RECO/STBYDB/ARCHIVELOG/2019_08_15/thread_1_seq_738281.18422.1016344287
RFS[5]: Opened log for thread 1 sequence 738480 dbid 1234567890 branch 948642470 
RFS[3]: Opened log for thread 1 sequence 738479 dbid 1234567890 branch 948642470 
RFS[4]: Opened log for thread 1 sequence 738478 dbid 1234567890 branch 948642470


Starting from Oracle 12c(12.1.0.2), controlfile transactions are enhanced. We are no longer request a controlfile enqueue before scanning a particular controlfile section. This is done to minimize locking issues and improve performance and scalability. Due to this change, the process reading a controlfile section may sometimes encounter ORA-235 if the same section is being modified by another process. When this occurs, the reading process will simply re-read the section again. You can safely ignore this message, it's not an issue. ----- EXPECTED BEHAVIOR in 12.1.0.2

The errors cannot be suppressed at database level by setting any event or initialization parameter. The reason is that although the errors are harmless, they do give an idea of the concurrency on the controlfile and can help debug issues relating to slow controlfile i/o etc so, this enhancement has been added specifically by development. It's best to modify any monitoring scripts to ignore/exclude this error. Since this is working as designed and these are simply informational, the messages can simply be ignored.




Hope it helps, subscribe to this blog to stay updated on latest Oracle Technologies and new articles.

Twitter : @rajsoft8899
Linkedin : https://www.linkedin.com/in/raj-kumar-kushwaha-5a289219/


No comments:

Post a Comment