Streams Apply Servers self-deadlocking

Dear list,

this is quite a specific question about streams ....

My source database produces too many LCRs for the destination database to keep up with (flow control kicking in, capture-to-apply lag is quite high). So I thought I'd increase the parallelism of my apply but this causes my a big headache: after 15 mins at most, the apply process aborts (abort on error is true) because of a deadlock caused by my 2 streams apply servers (AS03 and AS06). Checking dba_apply_error clearly indicates an ORA-60 as the cause for the apply to abort.

It's not a severe problem as I can easily rerun the transaction and restart the apply but still I'd rather solve this problem.

My question: did anybody experience the same problem with 2 parallel apply servers? If so, how did you fix it?

Facts:

Source and destination run on 32bit RHEL5u1, destination database is 11.1.0.6

Thanks for any suggestions!

Martin

--
http://www.freelists.org/webpage/oracle-l