What's wrong with Oracle 11g? I used a logon trigger to enable tracing and then I found the following in the tracefile. It looks to me like SQL*Net-Output among the database calls.
END OF STMT
PARSE #4:c=10000,e=11433,p=0,cr=0,cu=0,mis=1,r=0,dep=0,og=1,plh=468217235,tim=5331274397404
EXEC #4:c=0,e=0,p=0,cr=0,cu=0,mis=0,r=0,dep=0,og=1,plh=468217235,tim=5331274404320
2009-07-06 18:40:39.525417 : nioqsn:entry
2009-07-06 18:40:39.526123 : nioqsn:exit
WAIT #4: nam='SQL*Net message to client' ela= 1389 driver id=1413697536 #bytes=1 p3=0 obj#=62 tim=5459224997417
2009-07-06 18:40:39.527577 : nioqts:exit
2009-07-06 18:40:39.528352 : nioqts:exit
2009-07-06 18:40:39.529878 : nsmore2recv:testing for MOREDATA from transport
2009-07-06 18:40:39.530589 : sntpoltsts:exit
2009-07-06 18:40:39.531257 : nioqts:exit
2009-07-06 18:40:39.531946 : nioqts:exit
2009-07-06 18:40:39.532633 : nioqts:exit
2009-07-06 18:40:39.533310 : nsmore2recv:testing for MOREDATA from transport
BTW: After 9 minutes I killed the session and until then I didn't find the corresponding "SQL*Net message from client" to the "SQL*Net message from client" which started at tim=5459224997417. So I think we have a network problem... but why are the other strange lines in my beloved tracefile?