ROBLEM: ARCHIVE process get hang on Primary Database
The problem has been that the standby hardware have some faulty backpane.
This has caused processes in the standby to go into what is called as “D”
state where the processes gets stuck in a kernel call. So the RFS processes
on the standby are getting hung in the Operating System. This in turn causes the ARCH processes on the primary to hang. These ARCH processes on the primary are in actuality archiving the ORLs to the standby and to the local destination. So when they hang, they are also preventing the archival of the ORLs to the local destination. When enough of these ARCH processes hang, the primary runs out of ORLs and thus the primary itself hangs even though log_archive_local_first is set to true (default).
Manually transfer archives and apply them to the standby database daily.
bug 6113783 and bug 6987510