There is a Rman process that delete archives applied to a DataGuard instance. Has begun to appear this error, always with the last one sent you the parent instance. This never happened before and had been going on since a couple of days, when the unix box, primary and the standby went shutdown. Of course, I could occur in a point in time, when I run cron erasing some of the archives could be an archive just arrived. but this is controlled in this case only archives are generated in primary in a given time and only the archives are deleted on standby at another time than the first, which should reduce the possibility of collision to zero.
Researching My Oracle Support (formerly Metalink), I’ve only seen the note 1369630.1, ‘APPLIED’-Column not updated if Heartbeat-ARCH hangs, is what more similar to my problem, but it’s not the same, besides the database version was 126.96.36.199, so it would be out of this BUG
‘APPLIED’-Column not updated if Heartbeat hangs-ARCH (Doc ID 1369630.1)
Executing the following select
SELECT NAME, 'Applied:' || to_char (next_time, 'dd-Mon-yy:hh24:mi:ss') Time FROM V$ARCHIVED_LOG
It indicates that the archive is applied. So what is happening?
Examining the last bounce, I see is in real-time recovery, this is done with the command;
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE USING CURRENT LOGFILE;
Looking at the documentation (V10.2);
By default, the redo data is applied from archived redo log files. When performing Redo Apply, a physical standby database can use the real-time apply feature to apply redo from the standby Directly redo log files are being written Peak Gust by the RFS process.
After see “they are being written”, I think that as it will receive will apply redo vectors, maybe it could not know if it’s really the end (???) or may have to write more or RFS process had some pointer to archive file … The archive appears really like applied in the V$ARCHIVED_LOG, but RMAN does not delete because the think is necessary. To check this reasoning I stop the recovery mode in real-time and start the classic, using the command;
recovery mode in real time and get the classic, using the command;
This should be applied when the archive is “complete”. I run the cron and it works perfectly.
HTH – Antonio NAVARRO