TCPRC (Messages and Codes)

TCPRC000I

DATA RECORD ACTIVATED

Explanation: The token iiiiiiii identified by this message is not valid.

User Response: No action is required.

TCPRC009I

TCP/IP RECORDER SHUTDOWN ACKNOWLEDGED

Explanation: The Global Record TCP/IP data recorder is terminating.

User Response: No action is required.

TCPRC010I

TCP/IP RECORDER SHUTDOWN COMPLETE

Explanation: The Global Record TCP/IP data recorder has terminated.

User Response: No action is required.

TCPRC100W

HVCOMMON BUFFERS NOT LARGE ENOUGH, XXXXXXXX RECORDS LOST SO FAR

Explanation: There were more TCP or MQ events than the global capture buffers could handle. XXXXXXXX indicates the number of TCP/IP events lost.

System Action: Global Record presents this message when the first TCP/IP event is lost and again after each 100 TCP/IP events are lost. It also presents message TCPRC0915I to aid you with troubleshooting.

User Response: If this message occurs only at Global Record startup, and no TCP or MQ Global Recording Requests are active, ignore this message. Otherwise, complete the following steps.

• Ensure the started task dispatch priority is above all TCP/IP and MQ generating address spaces. If WLM goal mode is active, dispatch rate is determined by the VELOCITY of the performance group. The IBM default group, SYSSTC, is a good starting point for most systems.
• Access the Statistics panel from the Product Menu panel by issuing the primary command STATS and then access the TCP/IP specific panel by issuing the command TCP on the STATS primary command line. Evaluate the HVCOMMON Max Used and the Volume and Count fields and set parameters TCP_MQ_ECSA_BUFFER_SIZE (TCPBUFL) and TCP_MQ_ECSA_BUFFER_COUNT (TCPBUFS) accordingly. If activity varies throughout the day, turn on statistics by issuing the MVS MODIFY command “STATSON” for the Global Record address space. Monitor the HVCOMMON and Volume and Count fields during the increased activity periods and increase TCP_MQ_ECSA_BUFFER_COUNT and TCP_MQ_ECSA_BUFFER_SIZE to create additional global capture buffers to handle the activity spikes. See Performance Test Advanced Configuration Guide for more information.

TCPRC101E

nnnnnnnn RECORDS TOO LONG. INCREASE TCP_MQ_ECSA_BUFFER_COUNT*TCP_MQ_ECSA_BUFFER_SIZE VALUE

Explanation: Global Record has rejected nnnnnnnn records because they exceeded the size specified in either the MAXIMUM_TCP_RECORD_SIZE (TCPRECLM) SYSIN parameter, or TCP_MQ_ECSA_BUFFER_COUNT*TCP_MQ_ECSA_BUFFER_SIZE (TCPBUFS*TCPBUFL) if MAXIMUM_TCP_RECORD_SIZE (TCPRECLM) is not specified. TCP_MQ_ECSA_BUFFER_COUNT is the number of Global Record capture buffers to allocate for TCPIP and MQ, and TCP_MQ_ECSA_BUFFER_SIZE is each Global Record capture buffer size in kilobytes. The MQ messages have been discarded.

User Response: If you need to capture larger records specify a larger value in MAXIMUM_TCP_RECORD_SIZE (TCPRECLM). (See Performance Test Advanced Configuration Guide) Otherwise, this message is informational and can be ignored. If MAXIMUM_TCP_RECORD_SIZE is specified, it should be less than TCP_MQ_ECSA_BUFFER_COUNT*TCP_MQ_ECSA_BUFFER_SIZE (TCPBUFS*TCPBUFL) unless BUFFER_LATCH_WAIT=YES is used. With BUFFER_LATCH_WAIT=YES, the MAXIMUM_TCP_RECORD_SIZE can be up to 2,147,483,647. If MQ data capture is specified, BUFFER_LATCH_WAIT=YES will be enforced if PTF QFS200A is applied.

TCPRC102E

nnnnnnnn CPOOL ERRORS. INCREASE MAXIMUM_CONCURRENT_TCP_MQ_CALLS VALUE

Explanation: Global Record for TCP and/or MQ message process has rejected nnnnnnnn times. This happens when there is more concurrent message processing than the size specified in MAXIMUM_CONCURRENT_TCP_MQ_CALLS VALUE. At the first error message, an SVCDUMP named “ Hiperstation TCPRCDAT/TCP CPOOL PROBLEM” will be created. On the second occurrence, the Global Record task will be terminated (shutdown).

User Response: If the value for MAXIMUM_CONCURRENT_TCP_MQ_CALLS VALUE has been set below the default value of 100, specify a higher value and rerun. Otherwise, collect all messages in the system log, save the dump, and contact BMC Support.

TCPRC103I

JOBNAME: xxxxxxxx LENGTH: nnnnnnnn FROM:MQS

Explanation: This message is always preceded by message TCPRC101E. It provides JOBNAME and the size of MQ message for the last discarded one. For example: TCPRC103I JOBNAME: $MQPUT1Y LENGTH: 02001060 FROM:MQS.

User Response: No action is required.. However, see the User Response for the preceding message TCPRC101E.

TCPRC106W

HVCOMMON BUFFERS NOT LARGE ENOUGH, XXXXXXXX MQ RECORDS LOST SO FAR

Explanation: There were more TCP and MQ events than the global capture buffers could handle. XXXXXXXX indicates the number of MQ events lost.

System Action: Global Record presents this message when the first MQ event is lost and again after each 100 MQ events are lost. It also presents message TCPRC0915I to aid you with troubleshooting.

User Response: If this message occurs only at Global Record startup, and no TCP or MQ Global Recording Requests are active, ignore this message. Otherwise, complete the following steps.

• Ensure the started task dispatch priority is above all TCP/IP and MQ generating address spaces. If WLM goal mode is active, dispatch rate is determined by the VELOCITY of the performance group. The IBM default group, SYSSTC, is a good starting point for most systems.
• Access the Statistics panel from the Product Menu panel by issuing the primary command STATS and then access the TCP/IP specific panel by issuing the command TCP on the STATS primary command line. Evaluate the HVCOMMON Max Used and the Volume and Count fields and set parameters TCP_MQ_ECSA_BUFFER_SIZE (TCPBUFL) and TCP_MQ_ECSA_BUFFER_COUNT (TCPBUFS) accordingly. If activity varies throughout the day, turn on statistics by issuing the MVS MODIFY command “STATSON” for the Global Record address space. Monitor the HVCOMMON and Volume and Count fields during the increased activity periods and increase TCP_MQ_ECSA_BUFFER_COUNT and TCP_MQ_ECSA_BUFFER_SIZE to create additional global capture buffers to handle the activity spikes. See Performance Test Advanced Configuration Guide for more information.

TCPRC200W

RECORDING TERMINATED DUE TO FULL REPOSITORY

Explanation: A recording request has been terminated because the specified repository has been filled. A message will also be issued to the TSO user ID that started the request.

User Response: Use a larger repository if needed.

TCPRC201E

RECORDING TERMINATED DUE TO REPOSITORY I/O ERROR

Explanation: A recording request has been terminated because the specified repository has received an I/O error. A message will also be issued to the TSO user ID that started the request.

User Response: Check the system log for more messages that may indicate the specific I/O error.

TCPRC900E

TCP/IP RECORDER ESTAE ENTERED

Explanation: The Global Record TCP/IP data recorder has abended. Performance Test will terminate.

User Response: Collect all messages in the system log, save the dump and contact BMC Support.

TCPRC901E

NO RETRY, WILL NEED RESTART

Explanation: The Global Record TCP/IP data recorder has abended. Performance Test will terminate.

User Response: Collect all messages in the system log, save the dump and contact BMC Support.

TCPRC904E

BUFFER OVERRUN. PLEASE CALL  Hiperstation   SUPPORT

Explanation: An internal buffer error has occurred.

User Response: Collect all messages in the system log, save the dump and contact BMC Support.

TCPRC905E

NOT ENOUGH MEMORY FOR BUFFERS

Explanation: Global Record was unable to initialize the TCP/IP data recorder because there was not enough memory. Performance Test will terminate.

User Response: Increase the buffer size in the PARMLIB member.

TCPRC910I

MORE TADRS NEEDED

Explanation: The TADR data structure is used internally by Global Record to keep track of active TCP connections in the MVS system. The initial number of TADRs obtained is controlled by the MAXIMUM_TCP_CONNECTIONS (CONNCNT) parameter in the Global Recording startup parameters. This message appears if Global Record has to keep track of more TCP connections than are specified on the MAXIMUM_TCP_CONNECTIONS parameter.

System Action: Global Record tries to obtain additional TADRs.

User Response: Eliminate this message by setting a MAXIMUM_TCP_CONNECTIONS (CONNCNT) value larger than the maximum number of simultaneous TCP connections on your system. For each occurrence of this message, increase the existing MAXIMUM_TCP_CONNECTIONS value by minimally 25%. For example, if the existing MAXIMUM_TCP_CONNECTIONS is 1000 and you receive three of these messages, increase the MAXIMUM_TCP_CONNECTIONS to 1750. Setting an appropriate MAXIMUM_TCP_CONNECTIONS value improves performance.

TCPRC912I

MORE QADRS NEEDED

Explanation: The QADR data structure is used internally by Global Record to keep track of active MQ sessions in the MVS system. The initial number of QADRs obtained is controlled by the MAXIMUM_MQ_CONNECTIONS (MQSESCNT) parameter in the Global Recording startup parameters. This message appears if Global Record has to keep track of more MQ sessions than are specified on the MAXIMUM_MQ_CONNECTIONS parameter.

System Action: Global Record tries to obtain additional QADRs.

User Response: Eliminate this message by setting a MAXIMUM_MQ_CONNECTIONS (MQSESCNT) value larger than the maximum number of simultaneous MQ sessions on your system. For each occurrence of this message, increase the existing MAXIMUM_MQ_CONNECTIONS value by minimally 25%. For example, if the existing MAXIMUM_MQ_CONNECTIONS is 100 and you receive three of these messages, increase the MAXIMUM_MQ_CONNECTIONS to 175. Setting an appropriate MAXIMUM_MQ_CONNECTIONS value improves performance.

TCPRC915I

TCPRC916W

POSSIBLE LOSS OF TCP DATA. TSBF/TSBE DIAGNOSTIC INFORMATION FOLLOWS:

11901000 0000 E3E2C2C6 E3020FF8 11922000 118E0000 TSBFT..8.k...®.. 
11901010 0010 00000000 00000000 00020FD0 81000000 ...........}a... 
11901020 0020 00249A2E 00000000 E3E2C2D8 00020FD0 .. .....TSBQ...} 
11901030 0030 0015D080 C3C9C3E2 D6C2F0F2 1194472C ..}øCICSOB02.mâ. 
11901040 0040 918BE440 D7379DED 015F5349 000000BE j¤U P. é.^ ...Ø 
11901050 0050 14F1AC50 00000000 0000047C 00000000 .1Ú&.......@.... 
11901060 0060 00000000 00000000 00024220 00024674 ..........â...ãÈ

Explanation: Data loss may have occurred. The first 136 bytes of TSBF are displayed. The Free space (+14), Total space (+18), and Data filled (+24) are key information. If Data filled >= Total space - Free space, the message may be processed properly. In the sample above, there are several fields of interests:

• 
  
  • Total space = x'00020FD0' (135120 bytes)
  • Free space = x'00000000'
  • Data filled = x'00000000'
  • Data filled < Total space - Free space (0 < 135120)
  • +28 is TSBQ, not QSBE
  • +64 = x'00024220' is message size (148,000 bytes), therefore it should be segmented.

In other words, the job CICSOB02 did not fill the ECSA buffer yet, and it had been lost—that is, not processed.

This message can have a second part as follows:

POSSIBLE LOSS OF MQ DATA. SEGMENT 1 LOST

37F85008 0000 E3E2C2C6 E3020FF8 00000000 00000000 TSBFT..8........ 
37F85018 0010 37F96000 00000000 00020FD0 81000000 .9-........}a... 
37F85028 0020 0000001A 00000000 D8E2C2C5 00020FD0 ........QSBE...} 
37F85038 0030 4015D080 C3C9C3E2 D6C2F0F2 1194472C ..}øCICSOB02.mâ. 
37F85048 0040 008FE990 D762E33E 45953014 0002002F .±Z°PÂT.án...... 
37F85058 0050 387D4C18 00000000 0001147C 00000000 .'<........@.... 
37F85068 0060 00000000 00000000 00000000 00000000 ................ 
37F85078 0070 00000000 00000000 00024220 00024674 ..........â...ãÈ

The segment number is 2 bytes at +4C and total message size is x'00024220' (148,000 bytes) at +78.

User Response: Check whether the contents of the TSBF/TSBE or TSBE/QSBE display are of any concern. If they need to be analyzed further, collect the job’s log (the jobname is at offset +24), the Global Recording Request information, and the global recording task system log, then contact BMC Support.

TCPRC932I

LOCAL WSBF MAX X(xxxxxxxx).

Explanation:  Performance Test Global recording pre-allocates local VSBF for TCP/MQ, and the number of buffer allocated is X(xxxxxxxx). The number of buffers Performance Test allocates is approximately 25% of 31-bit address.

User Response: No action is required.

TCPRC933W

XXXXXXXX SPAN RECORD TOO LARGE.

Explanation: TCP or MQ message size is too large to handle. Message will be discarded, and the buffers (total number of buffers=XXXXXXXX) will be freed.

User Response: Increase HVCOMMON buffer allocation (maximum number of buffers=255, maximum buffer size=1024K).

TCPRC935W

TRIPLET COMPLETION QUEUE NOT EMPTY: TCPRTMCP VALUE X(nnnnnnnn)

Explanation: Triplet completion queue was not empty during global recording restart.

User Response: Review the previous Global Record task log for any erroneous shutdown occurrence. If you believe there is an product issue to be investigated, then contact BMC Support.