Oracle session creation takes long time

Question

Hi ,
I am getting the below mentioned error continously in the alert log file.
I am unable login through sqlplus on my remote database server from clients . Where as i am able connect locally on the server.
some times i am able to connect from clients(after 5 Mins). After that the session is working fine.
The response of tnsping goes to 20000 msecs.
O.S :LINUX rhel 5, oracle: 11.1.0.6 .
I have read some where that log files has reached 2GB. but in my case its not.
I have attached the listener log also.
Pls respond if u come across this error...................
ALERT LOG INFO:
Fatal NI connect error 12537, connecting to:
(LOCAL=NO)
VERSION INFORMATION:
TNS for Linux: Version 11.1.0.6.0 - Production
Oracle Bequeath NT Protocol Adapter for Linux: Version 11.1.0.6.0 - Production
TCP/IP NT Protocol Adapter for Linux: Version 11.1.0.6.0 - Production
Time: 31-AUG-2009 14:07:22
Tracing not turned on.
Tns error struct:
ns main err code: 12537
TNS-12537: TNS:connection closed
ns secondary err code: 12560
nt main err code: 0
nt secondary err code: 0
nt OS err code: 0
ORA-609 : opiodr aborting process unknown ospid (3654_1185936)
Mon Aug 31 14:47:11 2009
Fatal NI connect error 12547, connecting to:
(LOCAL=NO)
VERSION INFORMATION:
TNS for Linux: Version 11.1.0.6.0 - Production
Oracle Bequeath NT Protocol Adapter for Linux: Version 11.1.0.6.0 - Production
TCP/IP NT Protocol Adapter for Linux: Version 11.1.0.6.0 - Production
Time: 31-AUG-2009 14:47:11
Tracing not turned on.
Tns error struct:
ns main err code: 12547
TNS-12547: TNS:lost contact
ns secondary err code: 12560
nt main err code: 0
nt secondary err code: 0
nt OS err code: 0
ORA-609 : opiodr aborting process unknown ospid (5046_1214608)
30338,
LISTENER LOG INFO:
TNS-12518: TNS:listener could not hand off client connection
TNS-12547: TNS:lost contact
TNS-12560: TNS:protocol adapter error
TNS-00517: Lost contact
Linux Error: 32: Broken pipe
31-AUG-2009 18:24:00 * service_update * ORCL * 0
Mon Aug 31 18:24:20 2009
31-AUG-2009 18:24:20 * (CONNECT_DATA=(CID=(PROGRAM=)(HOST=__jdbc__)(USER=GP03751))(SERVICE_NAME=orcl)(CID=(PROGRAM=)(HOST=__jdbc__)(USER=GP03751))) * (ADDRESS=(PROTOCOL=tcp)(HOST=10.101.160.161)(PORT=2912)) * establish * orcl * 12518
TNS-12518: TNS:listener could not hand off client connection
TNS-12547: TNS:lost contact
TNS-12560: TNS:protocol adapter error
TNS-00517: Lost contact
Linux Error: 32: Broken pipe

Answer 1

Sathish11844964 wrote:
I am unable login through sqlplus on my remote database server from clients . Where as i am able connect locally on the server.When connecting locally, the OCI client driver starts up the dedicated Oracle server process to service that session (which is why +$ORACLE_HOME+ needs to be set so that the driver can find the oracle executable and why +$ORACLE_SID+ needs to be set so that executable know which instance's SGA to attach to).
When connect via TCP, the Listener accepts the connection. If it is a dedicated server request, the Listener does this part - creates a dedicated Oracle server process.
It is this part that is failing according to the details you have supplied.
First off - a work around if this is an urgent problem that needs to be resolved (and potentially also the solution for your underlying problem) - configure Oracle Shared Server and have the OLTP clients connect to shared servers.
some times i am able to connect from clients(after 5 Mins). After that the session is working fine.This sounds like a problem with dedicated Oracle server process taking an extraordinary long time to start up. This will usually be an o/s related problem, as it is the kernel's task to allocate resources and memory, load that executable from disk, initialise it, etc.
The response of tnsping goes to 20000 msecs.This is something different - it means that the Listener is slow to accept() (a socket call) an incoming connection request. Likely this is caused by the Listener waiting on a dedicated server processes to start, or insufficient server resources to accept and process the incoming (tnsping) connection.
I have read some where that log files has reached 2GB. but in my case its not.I would expect to see some kind of server problem. Number of processes allowed per o/s user being exceeded. Lots of swapping between the swap disks and memory. Perhaps loads of connection attempts, packet drops and so on at IP stack level.
TNS-12518: TNS:listener could not hand off client connection
TNS-12547: TNS:lost contact
TNS-12560: TNS:protocol adapter error
TNS-00517: Lost contact
Linux Error: 32: Broken pipeThis means that the Listener attempted to start an Oracle server executable, but the the executable did not respond to the Listener (inter process communication likely needs to happen). Thus the Listener could not hand over the client connection to the server process.
There are a couple of reasons why this can fail.. most of them I would think are o/s and resource related. (assuming that your listener.log is configured correctly)
I suggest checking the server and ensuring that it is capable of handling the current processing and memory loads on it. Also double check your listener.ora file to ensure that the broken pipes (lost communication) is not caused by errant parameter settings.