4 Replies Latest reply on Sep 17, 2010 2:57 PM by 796142

    Fatal NI connect error 12170

    796142
      I have a database version 11.2.0.1.0 Oracle 11 g cluster using two machines.

      I have 450 simultaneous connections

      my oracle cluster is Linux. my client machines are all windows xp

      I have had many complaints of slowness and make reading the alerts.log of my machines
      I realized that I had with the frequency following warning message:

      example :

      **Fatal NI connect error 12170.**

      **VERSION INFORMATION:**
      **     TNS for Linux: Version 11.2.0.1.0 - Production**
      **     Oracle Bequeath NT Protocol Adapter for Linux: Version 11.2.0.1.0 - Production**
      **     TCP/IP NT Protocol Adapter for Linux: Version 11.2.0.1.0 - Production**
      **Time: 02-JUL-2010 21:41:57**
      **Tracing not turned on.**
      **Tns error struct:**
      **ns main err code: 12535**

      **TNS-12535: TNS:operation timed out**
      **ns secondary err code: 12560**
      **nt main err code: 505**

      **TNS-00505: Operation timed out**
      **nt secondary err code: 110**
      **nt OS err code: 0**
      **Client address: (ADDRESS=(PROTOCOL=tcp)(HOST=192.100.9.55)(PORT=33384))**
      **Fri Jul 02 21:54:12 2010**
      **Setting Resource Manager plan DEFAULT_MAINTENANCE_PLAN via parameter**
      **Fri Jul 02 21:54:12 2010**
      **Starting background process VKRM**
      **Fri Jul 02 21:54:12 2010**
      **VKRM started with pid=36, OS id=19440**
      **Fri Jul 02 22:16:46 2010**

      despite the client machine to change the error is always the same:

      **Fatal NI connect error 12170.**
      **ns main err code: 12535**
      **ns secondary err code: 12560**
      **TNS-00505: Operation timed out**




      below is the default setting that I do in tnsnames.ora for all clients machines.

      really need help urgently


      MULTIMED_CLUSTER =
      *(DESCRIPTION =*
      *(ADDRESS_LIST =*
      *(ADDRESS = (PROTOCOL = TCP)(HOST = XXXX_cluster.XXX.local)(PORT = 1521))*
      *)*
      *(CONNECT_DATA =*
      *(SERVER = DEDICATED)*
      *(SERVICE_NAME = XXXXXX)*
      *(FAILOVER_MODE =*
      *(TYPE = SELECT)*
      *(METHOD = BASIC)*
      *(RETRIES = 5)*
      *(DELAY = 180)*
      *)*
      *)*
      *)*

      thank you

      Bruno Moreira
        • 1. Re: Fatal NI connect error 12170
          sb92075
          12170, 00000, "TNS:Connect timeout occurred"
          // *Cause:  The server shut down because connection establishment or
          // communication with a client failed to complete within the allotted time
          // interval. This may be a result of network or system delays; or this may
          // indicate that a malicious client is trying to cause a Denial of Service
          // attack on the server.
          // *Action: If the error occurred because of a slow network or system,
          // reconfigure one or all of the parameters SQLNET.INBOUND_CONNECT_TIMEOUT,
          // SQLNET.SEND_TIMEOUT, SQLNET.RECV_TIMEOUT in sqlnet.ora to larger values.
          // If a malicious client is suspected, use the address in sqlnet.log to
          // identify the source and restrict access. Note that logged addresses may
          // not be reliable as they can be forged (e.g. in TCP/IP).
          bcm@bcm-laptop:~$ oerr ora 12535
          12535, 00000, "TNS:operation timed out"
          // *Cause: The requested operation could not be completed within the time out
          // period.
          // *Action: Look at the documentation on the secondary errors for possible
          // remedy. See SQLNET.LOG to find secondary error if not provided explicitly.
          // Turn on tracing to gather more information.
          bcm@bcm-laptop:~$ oerr ora 12560
          12560, 00000, "TNS:protocol adapter error"
          // *Cause: A generic protocol adapter error occurred.
          // *Action: Check addresses used for proper protocol specification. Before
          // reporting this error, look at the error stack and check for lower level
          // transport errors. For further details, turn on tracing and reexecute the
          // operation. Turn off tracing when the operation is complete.
          I notice a pattern above.
          Site is afflicted with network gremlins.
          I expect that your Network folks won't embrace the follow,
          but I believe that Oracle is victim; not culprit.
          • 2. Re: Fatal NI connect error 12170
            damorgan
            By "cluster" do you mean RAC? If so state that explicitly.

            But for now runt this query.
            SELECT dbms_utility.current_instance, TO_CHAR(begin_time, 'HH:MI') time,
            60 * (select value from v$osstat where stat_name = 'NUM_CPUS') total,
            60 * (select value from v$parameter where name = 'cpu_count') db_total,
            SUM(cpu_consumed_time) / 1000 consumed,
            SUM(cpu_wait_time) / 1000 throttled
            FROM v$rsrcmgrmetric_history
            GROUP BY begin_time
            ORDER BY begin_time;
            Also check CPU utilization and cache fusion interconnect traffic.

            I have had a similar issue with an 11.1.0.7 and am far from enamored of RM.
            1 person found this helpful
            • 3. Re: Fatal NI connect error 12170
              796142
              Yes . I have a RAC ! I have a machine to be the storage. I have two machines as database.

              I reported this bug that is being described in the alerts database more than 50 times per day.

              In the beginning I thought it was a deadlock but the alert does not show me the deadlock occurrence of more than 45 days

              The query did not bring me any result row. ran the query connected with sys. what can it be?

              thank you very much for your help.

              Bruno Moreira
              Brazil
              • 4. Re: Fatal NI connect error 12170
                796142
                i dont have a solution.Thank to the help.