You can also get TNS 12540 as a result of too many processes. I would
guess what may be happening is all those aborted RMAN jobs are hanging
onto processes, hence everything is better after bouncing db. Perhaps
you should try upping both the unix and init.ora processes settings to
push the problem back. No reason not to if you aren 't even close to
actual swapping. And take a close look at your listeners. I can 't
recall if Solaris has swapinfo -a. netstat -a |grep WAIT might be
interesting, too, there might be a very long timeout on the extraneous
connections.
Joel =
Garry=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=
=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0 =
=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0
http://www.garry.to=20
=A0
