Processing of IMS OTMA Trigger Errors

This section describes how to resolve two error conditions.

Buffer Overflow

If the Substation ES OTMA client interface receives a message from OTMA, an error condition results, such as a buffer overflow (identified by the message SXI4712E or SXI4703E). That overflow occurs because the size of either the SRB buffer or the Substation ES communications buffer is too small to accommodate the message. Subsequently, the Substation ES OTMA client returns a negative acknowledgement (NAK) to OTMA and issues an error message to the log. The invalid transaction is requeued to the special destination SXITRNAK.

To resolve that error situation, do one of the following items:

  • If the SRB buffer size causes the error (indicated by the SXI4712E message), change the BUFFLEN-SRB parameter to accommodate the size specified in the message.

    If the communication buffer size causes the error, (as depicted by the SXI4703E message), change the BUFFLEN-TRANS parameter to accommodate the size specified in the message.

    To activate the changes, recycle Substation ES so that it can handle the message with the increased size, and then issue the REL-TRNAK command when ready to reprocess the NAKd triggers. No loss of IMS queue buffers will occur.

  • Issue an IMS Dequeue command to empty the queue; all data in queue is then lost. You need not recycle Substation ES.

Network Failures

If an ESB connection that relates to IMS is lost or broken, IMS trigger processing stops. Inflight guaranteed triggers that have not been acknowledged might be NAKd, and there is also a very tiny window where inflight reliable triggers might be NAKd as well. All NAKd triggers are requeued to the special destination SXITRNAK. When the ESB connection is reestablished, trigger processing restarts. NAKd triggers, however, are not automatically processed. To process them, issue the SET,IMS,REL-TRNAK command.

Guaranteed Trigger NAK Tpipe

In the unlikely scenario where a guaranteed trigger message is NAKd (failed delivery), the message is queued back in IMS to a Tpipe named SXITRNAK. To reprocess NAKd triggers, issue the SET,IMS,REL-TRNAK command. If an error occurs again when processing NAKd triggers, the current message is NAKd again and the REL-TRNAK processing is halted.