OEM after import sysman schema....

After the export/import of the sysman schema, we recoverd almost 23G of disk space. But yesterday the /oem partition again was 100% full ????
Analyzing the different diectories showed an udump directory with 23G of trace files !!!
So, something is wrong with OEM. I also noticed no notifications were being send to me. The trace files showed me an ORA-00600 [kwqidrdq: loop], [0], [0], [0], [0], [], [], [] error. Could this have something to do with the export/import ??
I check the repository for invalid sysman objects. No invalid objects. I bounced the database and after bouncing the database some emails were send to me with OEM notifications, but still a lot off notifications were waiting to be send.
Using the emdiag tool...

[oemprd01@amssys07 bin]$ ./repvfy verify notifications -level 9
Please enter the SYSMAN password:
---------------------------------------------------------------------------
EMDIAG: 2008.0321 Repository: 10.2.0.3.0 19-Jun-2008 07:47:08
---------------------------------------------------------------------------verifyNOTIFICATIONS
100. Notification backlog: 2
103. Stuck notification waiting to be delivered: 1090

I decided to tell Oracle about this and ask them if this could have anything to do with the wxport/import.
Still no answer, and my udump is filling up every day.