I had my database streams capture process stuck at WAITING FOR DICTIONARY REDO : FIRST SCN 52170068520 status when re-starting the capture process after a network un-planned downtime.

Using the script to check the minimum archived log required to start capture process, I am actually getting an SCN way later than the one reported in the status and I even made sure that the required archived log is available in the disk.

A quick search in Oracle Support got me to Doc ID 1487470.1

After Capture abort restart results in WAITING FOR DICTIONARY REDO: FIRST SCN instead of starting at the REQUIRED_CHECKPOINT_SCN (Doc ID 1487470.1)

So what the workaround suggested is to alter the first scn to the required checkpoint scn reported in DBA_CAPTURE, IF your the required_checkpoint_scn is higher than the first scn reported in the status (THIS IS IMPORTANT!)

Let's check my required_checkpoint_scn Stop the capture and alter the…