I think I have hit bug #4061041 it is on metalink. It is for 10.1.0.3.0 on Linux x86_64. What you see is a billion trace files in your udump and bdump dirs. I literally have 500+ trace files accumlate it the last 24 hours. They all have the text in them of:
before in kgllkdl(), hd= hex number
And it comes with a "severe loss of service"
Has anybody seen this? I checked on OTN and did see a patch to 10.1.0.4.0 on Linux x86_64.
Does anybody know where I can get 10.1.0.2.0 EE or a patch to 10.1.0.04.0?
When is 10gRel2 coming out?
thanks.
>On 6/23/05, Joe Smith <joe_dba@(protected)> wrote: > > Is this right or has something changed in 10g (10.1.0.3.0). This is on > > RHEL4 x86_64. > > > > I am consisting getting an: > > > > ora-12535 (See ora-12535.ora-code.com) tns: operation timed out > > > > when trying to connect from my client windows machine to the Linux >server > > with 10g Ent ed. > >I've been seeing alot of that lately. >Is there NAT involved between the client and server? >Does the oracle server have multiple network interfaces? > >Typically, a 12535 occurs after the redirect is sent to the client. >The client can tnsping the listener, but never connects to the >dedicated server process that has been spawned for it. > >hth. > >Paul
