We’ve been randomly getting a lot of the following error sequence lately:
Error number <50664> occurred. There is no additional description.
Cannot close pipe, due to error .
The data flow <DF_FinRecordCategory_SAP> process could not communicate with the job process.
And then it dies. I haven’t been able to find anything online about this. Anyone know?
(14.0) 05-07-13 00:57:25 (W) (5672:9504) INZ-050664: |Dataflow DF_FinRecordType_SAP
Error number <50664> occurred. There is no additional description. Notify Customer Support. Message parameters are: <31> .
(14.0) 05-07-13 00:57:25 (E) (5672:9504) SYS-050605: |Dataflow DF_FinRecordType_SAP
Cannot close pipe, due to error .
(14.0) 05-07-13 00:57:25 (E) (5672:9504) RUN-050408: |Dataflow DF_FinRecordType_SAP
The data flow <DF_FinRecordType_SAP> process could not communicate with the job process. For details, see previously logged
error <50605>.
(14.0) 05-07-13 00:57:25 (E) (9704:4956) RUN-050406: |Session Job_Dims_Common_GL|Work flow WF_Load_Common_GL_DIMS_SAP|Work flow C_FinRecordType_SAP|Data flow DF_FinRecordType_SAP
Data flow <DF_FinRecordType_SAP> received a bad system message. Message text from the child process is < >. The process
executing data flow <DF_FinRecordType_SAP> has died abnormally. For NT, check errorlog.txt. For HPUX, check stack_trace.txt.
Also, notify Technical Support.
(14.0) 05-07-13 00:57:25 (9704:4956) RUN-050409: |Session Job_Dims_Common_GL|Work flow WF_Load_Common_GL_DIMS_SAP|Work flow C_FinRecordType_SAP
It sounds like a different error in which the Dataflow has spawned a process and it eventually loses the ability to communicate with that child process.
The usual solutions:
Disable the “Run as a separate process” options
Find a way to make the Dataflow run faster so it doesn’t take so long. Often this can be done by simplifying what the Dataflow does or perhaps splitting the Dataflow into multiple Dataflows.
Avoid operations that accumulate large amounts of data on the job server. This means you need to eliminate or push down to the database operations such as DISTINCT, ORDER BY, GROUP BY, etc.
Thanks, I’ll look at those options; I don’t think we have “run as separate process” set anywhere.
We are looking at upgrading to 4.0 SP3 Patch 3 (we’re on patch 0); we ran everything today with no failures but since this is sporadic that’s not conclusive.
I hope you are trying to fetch data from SAP and using some file transfer method. Sometimes, if it is a high record count extraction, by the time the abap flow is done execution and the output is written into .dat file, the process which started this abap is not longer available. There could be many reasons like loss of connectivity between app server and the sap app server, could something cause the process to stop. This type of issues are not encountered regularly and cant be re-created. Network admins will have to answer your queries.
Update: I submitted a ticket to SAP and they felt that it had something to do with our ABAP program calls to SAP and suggested we look at the SAP logs. We found corresponding errors that indicated the CPIC buffer was filling up, which is what is used to communicate with external systems, and subsequent connection errors to the data services server. However, we increased the buffer size but still got an error this weekend, and they couldn’t find any errors at that time. So still looking.