Skip to Content
0

HALM Transports, XSENGINE TImeout Settings and Web Dispatcher

Jan 02, 2017 at 09:54 PM

276

avatar image

Hi,

We are running ECC6 on HANA 1.00.097 and are using the HANA Analytics functionality to report directly off the database for operational reporting.

We have a DEV->QA->PRD landscape and are using the HANA Native Transport process with Change Recording turned on in the DEV system. We have been encountering issues where transports which involve a large number of objects have been failing. We have implemented the HTTPSERVER and JSVM settings mentioned in OSS Note 2073243 and have been able to get to the point that the large transports finish successfully.

We do however still have one issue that I would like to resolve. In the HALM web console when monitoring a transport we are finding that with long running transports after 10 minutes the notifications in the web console will return an error message "Internal Error". The transport does eventually finish with a successfully completed status in spite of this error notification.

I have combed through the HANA configuration files and believe that the source of the "Internal Error" may be from the Web Dispatcher. Unfortunately I do have not experience managing or configuring this service and my "working theory" is based upon digging through the help files.

In the webdispatcher.ini the setting Profile -> icm/server_port_0 has a value of "PROT=HTTP,PORT=80$(SAPSYSTEM),PROCTIMEOUT=600". The PROCTIMEOUT means that the HTTP requests processing timeout is 600 seconds or 10 minutes. This is the amount of time we have observed that elapses before the "Internal Error" notification occurs. My "working theory" is that once the HALM makes it's initial request for the transport to start the PROCTIMEOUT clock starts, and if the transport processing in the target system does not reply back that the transport succeeded or failed within the 600 seconds the Web Dispatcher drops the HTTP connection and the HALM logs this as an unknown error.

The interesting part is that even with the "internal error" notification if the HTTPSERVER and JSVM are large enough / long enough for the transport to finish the target system will report back the status of the transport, and the HALM job log will show the final status of the job.

What I would like to know is if the HANA Web Dispatcher's primary function is to support the XS Engine, or in a system with ECC on HANA does the HANA Web Dispatcher also provide functionality to the ECC application layer? If the HANA WD is only for the XS Engine then I would like to try and change the PROCTIMEOUT to match the XSEngine max_request_runtime value that we have set. It would seem to make sense for these settings to align as the transport job should report back its status with this runtime.

If the HANA WD also provides functionality to ECC we would be reluctant to change the PROCTIMEOUT this significantly in the event that it would affect functions such as EDI, web service calls, etc.

If anyone could provide any insight or information on this issue and the architecture of the related functions that would be much appreciated.

Thanks! Mel

10 |10000 characters needed characters left characters exceeded
* Please Login or Register to Answer, Follow or Comment.

1 Answer

Best Answer
avatar image
Former Member Jan 02, 2017 at 11:33 PM
0

Hi Me,

HANA WD only provide support for XS of HANA database,it will not affect ECC WD .

So you can change the value of parameter .

BR

Atul

Show 1 Share
10 |10000 characters needed characters left characters exceeded

Hi Atul,

Thank you for your response!

Mel

0