Chapter 52 MHS (OSI/Message Handling System) Messages

Cause  OSI/MHS or GIP detected an unrecoverable internal error. If the error involves a crash loop, the text “crash-loop detected” appears in the operator message.

Effect  The affected process, OSI/MHS or GIP, terminates abnormally.

Recovery  Retain a copy of the message, contact the Global Mission Critical Solution Center (GMCSC), and provide all relevant information as follows:

Be especially alert to major failures such as processor and disk failures before the event occurs. Other information that could be helpful to the Global NonStop Solution Center (GNSC) includes:

If your local operating procedures require contacting the Global Mission Critical Solution Center (GMCSC), supply your system number and the numbers and versions of all related products as well.

Cause  OSI/MHS or the GIP detected an internal warning. If the GIP detected the error, see message 102 for detailed cause information.

Effect  If the error occurs within OSI/MHS, processing usually continues. If the error occurs in the GPI, the GIP terminates. See message 102 for information on the effect of GPI errors, as defined by the error code. For more information on the GIP, see the Tandem OSI/MHS Gateway Programmatic Interface (GPI) Reference Manual.

Recovery  Error codes 7106 through 7123 are informational messages only; no corrective action is needed. For error codes 7100 through 7105, or any critical errors reported by OSI/MHS processes, contact the Global NonStop Solution Center (GNSC) and provide all relevant information as follows:

If your local operating procedures require contacting the Global Mission Critical Solution Center (GMCSC), supply your system number and the numbers and versions of all related products as well.

Cause  OSI/MHS was unable to create a process in the named backup processor.

Effect  The attempt to start the process is abandoned.

Recovery  Examine the error code value, then take the appropriate action depending on the error produced. If error-code is a PROCESS_CREATE_ error, see Appendix D. For more detailed information, including recovery actions for PROCESS_CREATE_ errors, see the Guardian Procedure Errors and Messages Manual.

Cause  This message has several possible causes depending on the information contained in the nonrouting-reason field. These are some possible causes:

Effect  The local OSI/MHS MTA is unable to transfer a message to an adjacent MTA, local user, or gateway.

Recovery  Take action based on the nonrouting reason. For example, change the definition of ROUTE or APPL objects, or examine and correct the message.

Add or alter routes if necessary to allow transfer of subsequent messages and responses.

Cause  OSI/MHS could not determine how to route a message to a particular recipient.

Effect  The message was not routed to the recipient.

Recovery  Determine which user should have received the message. Next, check the routing configuration for that specific recipient. This message occurs in conjunction with MHS operator message 4. Take action based on the nonrouting-reason in that message.

Cause  An attempt was made to decode a message; however, it was not sufficiently decodable to permit the requested action (such as sending a nondelivery report). This error can be caused by corrupted data in the PDU store or receipt of an invalid byte stream.

Effect  The requested action (such as sending a nondelivery report) is not done.

Recovery  Contact the Global NonStop Solution Center (GNSC) and provide all relevant information as follows:

the Global NonStop Solution Center (GNSC) can assist you with specialized diagnostic tools.

Cause  An OSI/MHS message was transferred.

Effect  The message was transferred to the recipient.

Recovery  Informational message only; no corrective action is needed.

Cause  An OSI/MHS message was transferred to a specific recipient. This message occurs in conjunction with MHS operator message 7.

Effect  The message is transferred to the recipient. If a report cannot be delivered, the message provides nondelivery reason and diagnostic codes.

Recovery  If the message, probe, or report is successfully delivered, this message is informational only; no corrective action is needed. If a report cannot be delivered, see the Tandem OSI/MHS Management Programming Manual for information on the X.411 Nondelivery Diagnostic Reasons and Nondelivery Diagnostic Codes.

Cause  A SPI configuration request to add, alter, or delete one of the OSI/MHS object types shown above occurred.

Effect  The process continues.

Recovery  Informational message only; no corrective action is needed.

Cause  A message was stored in or deleted from the message store or report store.

Effect  The process continues.

Recovery  Informational message only; no corrective action is needed.

Cause  Message delivery was deferred or canceled.

Effect  The process continues.

Recovery  Informational message only; no corrective action is needed.

Cause  A file-system error was encountered.

Effect  If the file is not critical to the process, the process continues.

Recovery  Check the file associated with the file-system error number. For a description of file-system errors, see Appendix B. For more detailed information, including recovery actions for file-system errors, see the Guardian Procedure Errors and Messages Manual.

Cause  An OSI/MHS process rejected an OPEN request from an unauthorized requestor.

Effect  The process terminates. MHS operator message 1 can also occur with this message.

Recovery  If the requestor is an application, check the OPEN request to see whether it specifies the intended file name. If the requester is not an application, look for a recent occurrence of MHS operator message 1.

Cause  The message either was longer than the buffer transmitted, was odd, or was smaller than the smallest value for a valid request image. The process has either destroyed the valid buffer or received an invalid buffer.

Effect  The request is rejected.

Recovery  Examine, correct, and resend the SPI message.

Cause  The request was rejected with file-system error 2, which indicates that the operation is not supported. A received control and inquiry message was too short. The process has either destroyed the valid buffer or received an invalid buffer.

Any OSI/MHS process except the MHS manager can generate this event message. The message is usually caused by a process from another subsystem. It can also be caused by an OSI/MHS process if an internal error occurs.

Effect  The command is not executed.

Recovery  Correct the message byte count and resend the SPI message.

Cause  The request or response was rejected because it was invalid. This probably indicates a problem in your application software or OSI/MHS software.

Effect  The command is not executed.

Recovery  Contact the Global NonStop Solution Center (GNSC) and provide all relevant information as follows:

If your local operating procedures require contacting the Global Mission Critical Solution Center (GMCSC), supply your system number and the numbers and versions of all related products as well.

Cause  The MHS manager unsuccessfully attempted to restart a failed GROUP object.

Effect  The GROUP object is set to the STOPPED state.

Recovery  Follow these steps:

Cause  A process other than the MHS manager failed.

Effect  All processes in the group are stopped, then the group is restarted automatically.

Recovery  Informational message only; no corrective action is needed.

Cause  An invalid password was received from a message transfer agent (MTA) or an APPL object.

Effect  The attempted bind fails.

Recovery  Investigate repeated failed bind attempts, then reconfigure the OSI/MHS subsystem to disable the object.

Cause  A problem occurred with an APPL object or with a message, as indicated by diagnostic-reason.

Effect  The message store cannot process the message.

Recovery  Inspect the X.400 message. If the message store is full, the message store SQL table size should be extended; otherwise, make corrections depending on the reason displayed.

Cause  Not enough disk space was allocated for the process’s extended memory, or the disk was too fragmented.

Effect  The process stops.

Recovery  If the MHS manager fails to restart the process, provide more disk swap space for the process or run the DCOMP utility to compress the disk. If you still need assistance, contact the Global NonStop Solution Center (GNSC) and provide all relevant information as follows:

If your local operating procedures require contacting the Global Mission Critical Solution Center (GMCSC), supply your system number and the numbers and versions of all related products as well.

Cause  Either a problem occurred in the interaction between OSI/MHS and OSI/AS, or the problem was solved.

Effect  If a configuration problem has blocked OSI/MHS from using the specified device, OSI/MHS might continue to attempt to use the device until the configuration is corrected. If the configuration problem has been resolved, OSI/MHS is no longer blocked from using the specified device.

Recovery  If the message indicates a problem, check the configuration for the OSI stack and for OSI/MHS. Look for an error or a mismatch, depending on the API error subcode displayed. If the subcode in the operator message is in the range 103-106, check the specified attribute in the definition of the message transfer agent object and the underlying OSI configuration. If the subcode is in the range 107-113, check the configuration and status of the underlying subsystem and its components.

If the problem is not caused by a mismatch between OSI/MHS and OSI/AS addresses or by the omission of OSI/AS addresses from the configuration, then check for a lower-layer connection problem, such as rejected packet-switch network call requests. To determine whether the problem is at a lower layer, run a trace of the reliable transfer service (RTS) processes and examine them using the PTrace utility. Use the statistics counters of the RTS processes to identify and verify which processes are having the connection problem. For more information about running traces and statistics, see the SCF Reference Manual for Tandem OSI/MHS.

If the message indicates that a problem has been solved, it is informational only; no corrective action is needed.

Cause  A resource problem occurred for the named device or for the OSI stack as a whole, or the resource problem was solved.

Effect  OSI/MHS might continue to attempt to use the device until the resource is available.

Recovery  If the message indicates that there is a problem, check the configuration of the OSI stack for errors specified in the error codes, then change the configuration to correct the error. Look for an error or a mismatch, depending on the API error subcode displayed.

If the subcode in the operator message is in the range 103-106, check the specified attribute in the definition of the message transfer agent object and the underlying OSI configuration. If the subcode is in the range 107-113, check the configuration and status of the underlying subsystem and its components.

If the problem is not caused by a mismatch between OSI/MHS and OSI/AS addresses or the omission of OSI/AS addresses from the configuration, then check for a lower-layer connection problem, such as rejected packet-switch network call requests. To determine whether the problem is at a lower layer, run a trace of the reliable transfer service (RTS) process and examine it using the PTrace utility. Use the statistics counters of the RTS processes to identify and verify which processes are having the connection problem. For more information about running traces and statistics, see the SCF Reference Manual for Tandem OSI/MHS.

If the message indicates that the resource problem has been solved, it is informational only; no corrective action is needed.

Cause  An OSI stack configuration problem was detected at startup of an OSI/MHS GROUP object.

Effect  The process stops. OSI/MHS does not retry the startup.

Recovery  Make sure that the device name specified in the message is correctly configured. Look for an error or a mismatch, depending on the API error subcode displayed. Restart the GROUP object.

If the subcode in the operator message is in the range 103-106, check the specified attribute in the definition of the message transfer agent (MTA) object and the underlying OSI configuration. If the subcode is in the range 107-113, check the configuration and status of the underlying subsystem and its components.

If the problem is not caused by a mismatch between OSI/MHS and OSI/AS addresses or the omission of OSI/AS addresses from the configuration, then check for a lower-layer connection problem, such as rejected packet-switch network call requests. To determine whether the problem is at a lower layer, run a trace of the reliable transfer service (RTS) processes and examine them using the PTrace utility. Use the statistics counters of the RTS processes to identify and verify which processes are having the connection problem. For more information about running traces and statistics, see the SCF Reference Manual for Tandem OSI/MHS.

Cause  A resource problem for the named device, or for the OSI stack as a whole, existed at the startup of an OSI/MHS GROUP object.

Effect  The process stops. OSI/MHS does not retry the startup.

Recovery  Make sure that the device name specified in the message is correctly configured. Restart the GROUP object.

Cause  An OSI/MHS process received an unexpected message. The calling process might or might not be a part of OSI/MHS.

Effect  Unpredictable and possibly erroneous behavior occurs.

Recovery  Keep a copy of the message, contact the Global Mission Critical Solution Center (GMCSC), and provide all relevant information as follows:

If your local operating procedures require contacting the Global Mission Critical Solution Center (GMCSC), supply your system number and the numbers and versions of all related products as well.

Cause  There may be a problem with the SQL database, or there may be an internal programming error.

Effect  The process terminates.

Recovery  If there is a problem with the SQL database, correct the SQL database. If the problem is an internal programming error, keep a copy of the message, contact the Global Mission Critical Solution Center (GMCSC), and provide all relevant information as follows:

If your local operating procedures require contacting the Global Mission Critical Solution Center (GMCSC), supply your system number and the numbers and versions of all related products as well.

Cause  The alternate collector to which OSI/MHS is sending accounting messages is either available or unavailable.

Effect  OSI/MHS tries several times to send an accounting message to the alternate collector. If all retries fail, the accounting message is discarded.

Recovery  If the alternate collector is unavailable, restart the alternate collector process.

If the alternate collector is available, this message is informational only; no corrective action is needed. The alternate collector has been restarted and is ready to store accounting messages.

Cause  A file-system error from an I/O operation occurred.

Effect  The process stops.

Recovery  In some cases, the MHS manager restarts the GROUP object to which the process belongs. If the manager’s automatic recovery actions do not succeed, check the file file-name and correct the file-system problem. For a description of file-system errors, see Appendix B. For more detailed information, including recovery actions for file-system errors, see the Guardian Procedure Errors and Messages Manual.

Cause  A closed user group (CUG) violation was detected. Here are some examples of CUG violations:

Effect  OSI/MHS generates a nondelivery report for the invalid recipients.

Recovery  Check the current definitions of the CUG or CUGMEMBER objects. If those definitions are too restrictive, add or alter attributes as needed. For example, if the specified recipient should be able to receive messages from members of a CUG, add that user to the CUG.

Cause  A closed user group (CUG) violation was detected. Here are some examples of CUG violations:

Effect  OSI/MHS generates a nondelivery report for the invalid recipient.

Recovery  Check the current definitions of the CUG or CUGMEMBER objects. If those definitions are too restrictive, add or alter attributes as needed. For example, if the specified recipient should be able to receive messages from members of a CUG, add that user to the CUG.

Cause  An attempt was made to use a message transfer agent (MTA) bind password server, but no password server has been specified in the configuration.

Effect  OSI/MHS cannot establish an association with the remote MTA; therefore, OSI/MHS cannot send messages to the remote MTA nor receive messages from the remote MTA.

Recovery  Specify a value for the MTA-BIND-PW-SVR attribute for the OSI/MHS SUBSYS object, or change the PASSW-SERVER attribute to OFF for the adjacent MTA object.

Cause  A mismatch probably occurred in the OSI/MHS password and the adjacent message transfer agent (MTA) password.

Effect  OSI/MHS fails to establish an outgoing association with an adjacent MTA or it rejects an incoming request for an association from that adjacent MTA.

Recovery  Check the OSI/MHS configuration for the adjacent MTA and make sure that the passwords match those that the MTA bind password server is expecting. Tandem does not provide the MTA bind password server. Check the documentation for the password server that you are using.

Cause  A duplicate message or report message protocol data unit (MPDU) was discarded by the message store.

Effect  Processing continues.

Recovery  Check the distribution lists and accounting messages to determine if a recipient was specified more than once in the original message.

Cause  A reliable transfer service (RTS) association with an adjacent message transfer agent (MTA) was either successful or unsuccessful. An unsuccessful association is usually caused by an inconsistent configuration.

Effect  Processing continues.

Recovery  If the RTS association is accepted, no recovery is required. If the RTS association is rejected, check the reason code and correct the problem as necessary. For information about troubleshooting interconnection problems with adjacent MTAs, see the Tandem OSI/MHS Configuration and Management Manual.

Cause  The successful transfer by the RTS of an application protocol data unit (APDU) to or from an adjacent message transfer agent occurred.

Effect  Processing continues.

Recovery  Informational message only; no corrective action is needed.

Cause  An association with an adjacent message transfer agent either ended normally or aborted.

Effect  The association is ended.

Recovery  Check the reason code for the RTS termination. If the RTS terminates normally, no recovery is required. If the RTS aborts repeatedly, use PTrace to discover the reason for the abort. For information on using PTrace, see the Tandem OSI/MHS Configuration and Management Manual.

Cause  A successful or unsuccessful association establishment with a remote user agent occurred.

Effect  Processing continues.

Recovery  If the remote operations service (ROS) process association is accepted, no corrective action is needed. If the RS process association is rejected, take any of the following actions to diagnose the problem:

Cause  The transfer of an APDU by the remote operations service (ROS) process to an RUA (outgoing transfer) or from an RUA (incoming transfer) occurred.

Effect  Processing continues.

Recovery  Informational message only; no corrective action is needed.

Cause  This accounting message indicates whether the remote operations service (ROS) association terminated normally.

Effect  Processing continues.

Recovery  Informational message only; no corrective action is needed.

Cause  Either a successful LO association (Accepted) or an unsuccessful LO association (Rejected) occurred.

Effect  Processing continues.

Recovery  If the association is accepted, this message is informational only; no corrective action is needed. If the association is not accepted, check the surrounding operator messages and take the recovery actions that they describe.

Cause  The transfer of an application protocol data unit (APDU) between a local operations (LO) process and a local user agent occurred.

Effect  Processing continues.

Recovery  Informational message only; no corrective action is needed.

Cause  The local operations (LO) association either terminated normally or aborted.

Effect  Processing continues.

Recovery  Informational message only; no corrective action is needed.

Cause  The OSI/MHS subsystem either routed an MPDU or determined that it could not be routed.

Effect  OSI/MHS queues the MPDU for delivery or relay.

Recovery  If the MPDU was routed to an APPL, a DLIST, or a ROUTE, no recovery is required. If the MPDU is labeled UNROUTEABLE, look for occurrences of MHS operator messages 4 and 5 near this message. Operator messages 4 and 5 provide more information about routing problems. For more information about unroutable MPDUs, see the Tandem OSI/MHS Configuration and Management Manual.

Cause  The message store (MS) discarded a duplicate message or report message protocol data unit (MPDU).

Effect  Processing continues.

Recovery  Review the distribution lists and accounting messages to determine if a recipient was specified more than once in the original message.

Cause  The operations performed by the message store user were counted. The count begins either at the start of the association or at the previous occurrence of this message.

Effect  Processing continues.

Recovery  Informational message only; no corrective action is needed.

Cause  An MS operation result was either successful or unsuccessful.

Effect  Processing continues.

Recovery  Informational message only; no corrective action is needed.

Cause  The usage level of a resource exceeded a threshold value.

Effect  In general, if a usage level exceeds a threshold, the incoming message or request is rejected. A message store process, however, behaves as follows:

Recovery  If the usage % level is above the threshold, you must reduce the usage of the resource to an acceptable level. You can reduce the usage of the resource in two ways:

If this message indicates that usage % is below the threshold, the problem has been corrected. The message is informational only; no correction is needed.

Cause  A message in a message store (mailbox) changed state in one of the following ways:

Effect  Processing continues.

Recovery  Informational message only; no corrective action is needed.

OSI/MHS message store (MS) processes can generate this event message.

If you use a version of MHSMGR that does not support the SCF product module that has Event 50, the command ALTER SUBSYS #subsystem‑name, ACCT‑RESET [ ON | OFF ] fails, and the message “invalid token received” is returned.

Cause  A gateway input queue is not being accessed in a timely manner by the Tandem Secure Internet Messaging System (TSIMS) MTA or any other gateway application.

	NOTE: After every twentieth message queued to the gateway input queue, the OSI/MHS MTA retrieves the time of enqueue for the first message on the queue. If the first enqueued message has remained on the queue longer than the maximum time configured in the gateway object, event 52 is generated. This event is also generated every time an additional 20 messages are enqueued, as long as the message has not been retrieved from the queue. After the message is retrieved, no additional event is generated.

Effect  None. Information only.

Recovery  Check the appropriate gateway application program or extend MAX‑TIME‑EXPIRED using the ADD or ALTER GATEWAY command (see the SCF Reference Manual for Tandem OSI/MHS for further information about the ADD GATEWAY command).

Note that this procedure only prevents the generation of this event. You may need to determine why the gateway application process did not remove messages from the gateway queue within the specified time.