Checkpoint is not complete when there are too many inserts in SYS_EXPORT_FULL_01 during expd
Hello everyone!
It happens that I am surprised with the amount of switch logfiles generated by the DB when an EXPDP is executed, in a minute, 2 or 3 archivelogs are generated during the export with datapump. I researched with logminer the contents of the archivelog and I see that all transactions are insert/update on the table SYS_EXPORT_FULL_01. In 10 minutes of expdp execution, 500,000 records were created in SYS_EXPORT_FULL_01. Obviously, with so many insert the redologs are filled more frequently and generate too many archivelogs. This has caused alertlog to present the following message:
Thread 1 cannot allocate new log, sequence 2351