cancel
Showing results for 
Search instead for 
Did you mean: 

Message flow in SAP PI 7.31 JAVA only for ABAP Proxy synchronous scenario with Oracle DB

Former Member
0 Kudos

Dear Experts,

I am working on a synchronous scenario Sender ABAP Proxy <===> SAP PI 7.31 JAVA only <===> Oracle Database.I have successfully

configured all the proxy configurations as well both Sender SOAP and JDBC receiver channels are showing no error when pinged from

Cc monitoring in NWA.

1.During testing at Tx SPROXY , I am triggering a message and after 5 minutes I receive an PARSING error message with 2 messages in ECC

local IE i.e. Tx SXMB_MONI.

Note : In NWA , I can see only 1 message that too in cancelled status.

2.Seeing the trace of the messages I assume that the messages have not moved from sender ECC system. I am still confused with flow of message

from the Message Audit log. Audit log of error messages.Could any one share the full Audit log of a successful message generated in SAP PI 7.31 JAVA only.

3.I observed that the QIN Scheduler status in ECC is in INACTIVE status. There is no queue id for 2 messages generated in ECC. Does

it should in ACTIVE status in Tx SMQR..

Regards

Rebecca

Accepted Solutions (1)

Accepted Solutions (1)

Former Member
0 Kudos

Hi Rebecca,

1.) your scenario is synchronous. The status of the request message (the grey dot) is "log version". Simply said thats only a logged version for displaying purposes. This is not a final message state.

2.) seems to me the request is taking too long (problaby on receiver side, oracle database?), so after 5 Minutes a timeout occurs. This corresponds to your error message ("Message expired").

Check the logs on receiver side for any clues whats taking so long.

I'll see if i can attach an audit log from our 7.31 machine.

3.) synchronous messages do not generate queue entries, you will never see them in smq1/2.

So the scheduler status should be irrelevant in this scenario.

Kind regards,
Thomas

Former Member
0 Kudos

Hi Thomas,

Valuable inputs..

1. Since I don't have any audit log for successful message, would you be able to share referring to the

below screenshot whether the message was put into receiver JDBC adapter and JDBC adapter tried to retrieve the Select response from Oracle Database.

2.I have done some setting in Advance mode of the receiver JDBC to increase the time out and other parameters as per SAP note but still the issue is same. If you perhaps have some experience in such scenarios, please share some input so that I can ask the Oracle Database administrator exactly what is required from our side.

3.Regarding the slow response from Oracle , there was some suggestion that the datatype of the request and response should match the column name of the Oracle database. I did that.. Do you have any more suggestion that may be hampering the generation of the response message from the Oracle like SAP note or document.

Regards

Rebecca

Former Member
0 Kudos

Hi Rebecca,

1.) synchronous scenarios can be tricky to debug. From your screenshot, I cannot say for sure wether the message reached the JDBC adapter.
You can try the following:

  • check communication channel monitoring for errors in your sending and receiving channel. Sometimes channel monitoring gives more details than message monitoring.
  • increase the log level for JDBC adapter and/or the audit log. To do that, goto NWA -> Troubleshooting -> Logs and Traces -> choose "Log configuration". Navigate to the categories:
    / Applications / ExchangeInfrastructure / AdapterFramework / Services / AuditLog
    and/or
    / Applications / ExchangeInfrastructure / AdapterFramework / Services / ADAPTER / ADMIN / JDBC
    and set the severity to your desired level.

      "Debug" should do the trick. Try "all" just to be sure, but bear in mind that a log level too high may cause performance issues by itself.

      This screenshot may help:

2.) You're still receiving a timeout after 5 minutes? I'm thinking the limiting factor here is the the timeout setting in ECC (sender-side). Check the ICM-settings in ECC, if the timeout has a lower value than your settings in the JDBC adapter, than this settings will have no effect.


3.) Sorry, my experience with JDBC is quite limited. I don't have any more suggestions, at least not without any more details, like an errormessage or something. But maybe increasing the log level can provide that.

Kind regards,
Thomas
 

Former Member
0 Kudos

Hi Thomas,

Many thanks for the input...

Since my scenario has been configured as ICO, I had already set the logs for staging and logging in ID as well as NWA but in vain.

Meanwhile please share the Audit log of a successful synchronous message as discussed earlier in the thread if possible.

Regards

Rebecca

Answers (0)