SAP Note 19851 - ISHCM communication (no longer) works

Symptom:

Communication using IS-HCM does not work. The transceiver on the non-SAP system side is not started.

Cause and prerequisites

Various causes are possible.
It may be that the local (changed) installation is incorrect.

Solution

Make sure to check the following facts (also see Implementation Guide (IMG) "SAP Healthcare - Industry-Specific Components for Hospitals -> Communication -> Appendices / Communication Setup Procedures -> Troubleshooting"):

1. Did the error only occur once? In this case the actual cause is most probably a temporary net(work) problem. Check to see whether the error occurs again.

If the error occurs repeatedly, the error must be identified in more detail (by means of the following actions or checks)

2. Carry out an installation check (program RNCCHK01). The results log should be analyzed in detail. If possible, correct all errors listed by means of corresponding Customizing. More information on this is given in the Implementation Guide.
In the case of a central installation (HCM transceiver on an R/3 application server or database server), make sure that all scripts, programs and data directories are outside of the R/3 path /usr/sap. Otherwise, you run the risk of simply losing these objects during an upgrade.

3. Have you archived a list of all IS-HCM settings for the partner system in question? This is to be recommended after an installation has been completed succesfully. Have settings been changed? Does communication work (again) when you undo the changes?

4. After an unsuccessful communication attempt, is there an entry (or several) with an "CPIC-RC" in the system log?
If no: Another error exists: Check program(s).
If yes: Double click on the respective line to determine the significance of "SAP-RC" also issued. Does this help?

5. Was the local installation successfully tested using the local test? The program should start under the Gateway ID! From within the Gateway host, the Gateway ID must always be able to reach the target computer by means of remsh.
(For NT also read Note 64825)
For Gateway problems:
In Transaction SMGW, increase the trace level (Goto -> Trace) and display the trace file after a new transfer attempt.

6. Have the shell scripts been changed? Are all paths correct? Is the Gateway user authorized to execute the script/program? Are the scripts contained in the path of the Gateway user?
Caution: "echo" commands in the scripts prevent their correct flow. File commands (move, copy, remove, ...) can also cause windows to hang.
In the event of problems, you can temporarily specify absolute paths in TXCOM to ensure that the transceiver files are found.
The Gateway user is the user under whose name the trace files are created. Under Unix, this can also mostly be determined by ps -ef | grep gwrd.

7. Is the transfer file correct (new lines, end of file; see Note 30195)?

8. Does a '.tmp' file still exist?
(Caution: Many operating systems no longer call this NTISH.tmp, but assign a technical number instead)
In this case, check whether the messages contained therein

a) were already imported by the partner system (with HCM dispatch). If not, simply rename the file NTISH.tmp (or NTHL7.tmp) as NTISH.dat (or NTHL7.dat). Otherwise delete the '.tmp' file.

b) were already included in IS-H (with HCM receipt). If not, maintain the R/3 table TNCRESPO (using Transaction SM31/SM16) by entering the corresponding TXCOM entry with the final digit '2' as SDEST, the file as NAME, and an 'F' (for 'FALSE') as QUITT. Finally, start program RNCCDERR. Afterwards the '.tmp' file should have disappeared. A '.dat' file (with possibly new records) should exist.

9. Are the background processes (still) running? Do the background processes report any errors when run (HCM monitor, check logs, also try online)?

10. SM51: Determination of instances
__
The computer must be the same in both columns. Also check the instance number (mostly 00). If you entered a Gateway name in TXCOM, the instance number there must match the instance number here.

11. Is the code page correct?
In TN02S (partner system table), check whether an entry was made for the code page. This should generally be empty. If, however, you swap data with an EBCDIC computer (AS/400), the code page 1103 must be defined therein.

Additional key words

RNCSND01, RNCSND02, RNCGETxx, RNCGET00, RNCGET01, RNCGET02, RNCGET03, RNCGET04, RNCGET05, communication module, PUT_TAB, GET_TAB, GET_RESP, SEND_TAB, dispatch, export, receipt, consulting