Database Software

HI Guys
I have an environment which is being replicated in 2 ways (DML,DDL ) RDBMS 10.2.0.4 , each RDBMS is running in RAC (3 nodes one RDBMS) and another running (2 nodes RAC) everything was running without problems for the last 3 months however last weekend I got

Sat Feb 14 23:18:41 2009
STREAMS Warning: ORA-26667 during checkpoint retention check <= this message is the first kind of message that we got
Sun Feb 15 05:18:43 2009
STREAMS Capture C 1: first scn changed.
scn: 0x0009.79468518

However I took a look that day and I dint find any issue I looked for that problem in metalink but I didn't find any problem , this message came out when the checkpoint retention time moved

the checkpoint retention time in streams is a number of days that the capture process retains checkpoints before purging them automatically . A capture process periodically computes the age of a checkpoint by subtracting the next_time of the archived redo log that corresponds to the checkpoint from dba_capture.first_time of the archived redo log file containing the required checkpoint scn
if the resulting value is grater than the checkpoint retention time then the capture process automatically purges the checkpoint by advancing the dba_capture.first_scn value all the archived retained by the dba_caprure process can be seen using dba_registered_archived_log view
the default retention time is 60 days

i didnt find any error in the RDBMS therefore I didnt do anything but on Monday morning I got a lot of replication errors caused by
ORA-01403 , after fixing all those errors I changes of the database that I got that error were not replicated across the other database , it could explain why many tables got out of sync , seems the reason was the propagation process was hanging , I stopped and restarted no changes were received in the other database , as a workaround I recreate it and the problem was fixed . Have you had a problem like this before ?

Rgds