I can PING the server where PROD resides. I don't have access to the SQL*Net documentation here to look up this error message. When I go to the site for looking up Oracle 9i messages it does not list this one. We are on Oracle 8.05.
Here's the info from MetaLink - Note that it applies mostly to Windows/NT though there is one point for Solaris. The formatting is really poor since I just did a cut and paste. To see it properly formatted you will access to MetaLink.
PURPOSE To provide an overview of how to verify and handle errors ORA-12151 and ORA-12571. SCOPE AND APPLICATION This notes applies to anyone facing intermittent SQL*Net read and write error when using Oracle SQL*Net or Net8 on Windows platforms. ================================================== ============================ -------------------------------------------- ORA-12151 and ORA-12571 errors on Windows NT -------------------------------------------- Intermittent SQL*Net TCP/IP read and write errors are sometimes encountered on Windows NT. The underlying reasons of these errors are a synchronization error in the TCP/IP layer on Windows NT. To help prevent this kind of error, a few things can be adjusted to help the synchronization:
1. TCP.NODELAY parameter This parameter may be added to the PROTOCOL.ORA file in the NETWORK\ADMIN directory. In most cases, TCP/IP data sent across the network is buffered until at least a complete network packet can be send. This means that in certain cases, commands are not issued directly, and kept buffered until some other data can be sent as well. This has the potential to generate time-outs and errors. To avoid this, the delay can be switched off. tcp.nodelay = yes
2. DISABLE_OOB parameter Another possible cause of ORA-3113/ORA-12151 is caused by a known issue affecting the TCP/IP stack on Sun Solaris, for which the only available workaround is to disable out-of-band breaks. The issue is discussed in detail in [NOTE:1068560.6]. [NOTE:1016295.4] and [NOTE:120498.1] discuss this little known and used parameter.
2. Disabling AUTOMATIC_IPC On client PC's, checking for IPC connections is pointless as there is usually no database residing on them. To save time during the connection phase, set AUTOMATIC_IPC=OFF in the SQLNET.ORA file.
3. NAMES.DIRECTORY_PATH to force use of TNSNAMES and/or ONAMES If you have a static environment, it is recommended to explicitly specify this parameter in the SQLNET.ORA file. The parameter specifies how the Transparent Network Substrate (TNS) resolution is to take place. By default, if this parameter is not present - the SQL*Net layer will first check if there is a Names Server anywhere on the network, after which it checks for the existance of local TNSNAMES.ORA file. If you only have a TNSNAMES.ORA file, it is recommended to explicitly specify the parameter to avoid unecessarily searching for Names Servers - this not only speeds up TNS resolution, but also prevents unecessary SQL*Net trace file generation when SQL*Net tracing is enabled. The parameter value is a comma separated list, with the possible values of: TNSNAMES (TNSNAMES.ORA), ONAMES (Oracle Name Server) and HOSTNAME (Directory Cell Environment (DCE)).
4. TCP/IP timeouts on NT The default retransmission count on Windows NT is 5, before it detects that the network is down. With the value of 5, the actual timeout is aproximately 15 seconds. This default value can be easily increased to a higher value by modifying TCP parameters in the Windows registry i.e. HKEY_LOCAL_MACHINE System CurrentControlSet Services TCP/IP Parameters TcpMaxDataRetransmissions REG_DWORD "number" By default, the parameter is not present in the registry. If modifying the parameter for the first time, it will need to added. The parameter can be useful on both client and data server. The recommended first course of action is to add the parameter on the machine generating the SQL*Net errors. If problems persist, add or modify the parameter in the registry of the data server or other machine/s.
5. TCP/IP keepalive on NT KEEPALIVE is an extension to TCP/IP which enables the closing of dead connections that are no longer being used. Problems can occur when the server does not close a connection after a client process has disappeared or terminated abnormally. This typically happens when a user switches off or reboots their machine whilst still connected to Oracle. Note: this is not an Oracle problem, but a limitation of TCP/IP, which has no way of knowing whether a remote connection has disappeared. This feature is enabled by default on Windows NT, however the deafult value is 2 hours. Problems can arise however if the timeout value is set too low for some heavily used or slow networks. Under these conditions, the KEEPALIVE registry value can be used to specify a KEEPALIVE value before a connection gets cut. HKEY_LOCAL_MACHINE System CurrentControlSet Services TCP/IP Parameters KeepAlive REG_DWORD "number" A value of 10 minutes is a typical value used. Again, the parameter can be useful on both client and server. Start with the machine generating the error, and if needed, add it to the data server or other machine/s.
6. TCP/IP timeouts on Windows 95/98 The same parameter may also be used under Windows 95. It performs the same functionality, however only the location of the parameter is different. HKEY_LOCAL_MACHINE System CurrentControlSet Services Winsock Parameters TcpMaxDataRetransmissions REG_DWORD "number" Again, the parameter is not present in the registry by default. This means the parameter must be added to the registry the first time it is modified.
7. SDU & TDU parameters Part of the problem may be the sequence of information that is transmitted. If there are disruptions in the sequence, errors ORA-12151 and ORA-12571 can also appear, alerting the application that not all information has been sent across the network succesfully. The sequence of information is determined by the amount of data the program is sending and the actual size the protocol can send across the network at a time. The more data the program wants to send in one 'go', the more sequences and transport packets will have to be made. By default, SQL*Net uses a Session Data Unit SDU) of 2048 bytes (2Kb) and a Transport Data Unit (TDU) of 32768 (32Kb) bytes. On standard Ethernet connections, without modification, the SDU is 1500 bytes and TDU 8760 bytes. With these values, each data request made by SQL*Net must be split into several smaller packets to be able to be transmitted. Therefore, where errors occur, it is recommended to minimise the creation of unecessary additional packets by synchronising the SDU and TDU parameters at the SQL*Net level with those of the actual network topology/protocol in use. To use non-default SDU/TDU values, the parameters must be configured within both client and server SQL*Net configuration files as follows: TNSNAMES.ORA: ------------- ORCL.WORLD = (DESCRIPTION = (SDU=1500) (TDU=8760) (ADDRESS_LIST = (ADDRESS =(PROTOCOL=TCP)(HOST=foobar)(PORT=1521)) ) (CONNECT_DATA = (SID = ORCL) ) ) LISTENER.ORA: ------------- ... SID_DESC_LISTENER = (SID_LIST = (SID_DESC = (SDU = 1500) (TDU = 8760) (SID_NAME = ORCL) ) ) For additional information regarding SDU and TDU parameters, refer to [NOTE:44694.1]: SQL*Net Packet Sizes (SDU & TDU Parameters).
8. Setting a new TDU size on Windows NT You can modify the TDU size on Windows NT, via the TcpWindowSize parameter: HKEY_LOCAL_MACHINE System CurrentControlSet Services Tcpip Parameters TcpWindowSize REG_DWORD "number" Additional information about Windows NT network parameters: ----------------------------------------------------------- Q120642: TCP/IP & NBT Configuration Parameters for Windows NT http://support.microsoft.com/support.../Q120/6/42.asp Q140375: Default MTU Size for Different Network Topology http://support.microsoft.com/support.../Q140/3/75.asp