Hello,
We recently encountered an issue in our SNC system. Following is the sequence of events:
- A ProductForecastRevisionNotification_In message arrived from a B2B partner and was successfully posted to SNC using the standard proxy adapter (in namespace http://sap.com/xi/SNC/Global). This message was processed on one of several app servers (call it App Server A) tied to the same SNC instance/System ID.
- 3 minutes later we triggered a custom program which extracts the same data which was previously posted to SNC using Function Module /SCA/TSDM_TS_GET. This program ran on a different app server (call it App Server B).
- This extract did not include any of the updates from the message that arrived 3 minutes earlier. The same program was run a few hours later and fetched the data as expected.
The PFRN_In message was relatively large: ~14MB. But it's our understanding that all DB updates should be committed before reporting the message as 'Completed Successfully' in sxmb_moni. So we would not expect to see any lag time between completion of message posting and availability of data for extraction. We have never seen behavior like this before, so we're a bit stumped on this. We're thinking that perhaps some time period is necessary to ensure consistency across all of the app servers tied to the instance, but really not sure about this.
We would greatly appreciate any helpful suggestions about how to approach this problem, or any comments from anyone who has had an experience like this.
Thanks,
Devin
