please provide some version information about IS, SAP Adapter and JCo Libraries.
Please check your IS configuration for proper java memory settings (you might need to increase these values to avoid swapping or OutOfMemory exceptions).
You might want to rethink your implementation approach.
Instead of directly processing the large volume just publish the splitted data to a Messaging provider, i.e. UM or Broker, and then listen to this messaging queue and process the data in smaller chunks.
Yes…Also please consider async approach using pub/sub to process that kind of large volumes records and work with your SAP team as well to send the records as in batches 100 or approx. per each run to the adapter from processing/performance point of views.
We have agreed with our SAP Partners that there should not be more than 100 IDocs in one master IDocList when sending IDoc to us.
This is due to the fact that our IDocs can be quite large (about 1.500 fields which will be 150.000 fields in total plus EDI_DC40 control header).
Description webMethods SAP Adapter Adapter Version - 7.1.0.16.906 Updates
SAP_7.1_Fix14
SAP_7.1_Fix15
SAP_7.1_Fix16 JCo Version 3.0.14 (2015-12-16) JCo Middleware JavaRfc JCo Middleware Version 2.2.17 JCo Native Library Version 721.619
After doing some R&D , we understand like we cannot control the incoming records from SAP through RFC in webMethods.
So we followed batch process, where SAP will divide 5Lakh records from their Master tables, into multiple files in a folder and webMethods will pick and process.