Problem with older Enterprise 4.11 SAP ALE

Is anyone out there using the older webMethods Enterprise 4.1.1 Broker besides us? I have very stable EDI and EAI integrations in place that I was hoping to get at least 1 more year of service from. Unfortunately, a division of the company is implementing the latest and greatest SAP ERP 6, Unicode.
I tried to connect an Enterprise ALE Adapter version 4.1.1 to the new SAP but the adapter JVM aborts as soon as it starts to commuinicate with SAP. The error window in Adapter_config on a test run has the following error.
crfcglob.cpp:480: failed assertion ‘nLength == 0 || nLength == nTrueLen’
SIGABRT 6 Abort (generated by abort(3) routine)
si_signo [6]: ABRT
si_errno [0]
si_code [-1]: SI_LWP [pid:14139, uid:11431]
stackpointer = FFBED018

Nothing gets written to the dev_rfc.trc file for webMethods, but the SAP gateway monitor knows about the address of my adapter, and fails trying to read from that IP:port (reading a response to something it just sent I suspect)
Anyone got any ideas if there is anyway to get WM Enterprise 4.11 broker ALE to talk with newer releases of SAP? Otherwise what is the quickest path to link up my B2BServer (vers. 401) with all my custom maps to the new SAP system. Currently the Enterprise integrations (ATC based) convert the IDOC events into the OAGIS XML document standards that pass into IS through the bridge adapter. In the IS server, unique customer, vendor, and banking maps convert the data to a multitude of X12 formats, audit the various version in an Oracle database with display routines for viewing this log and each transaction in pretty formats. Replacing everything is the long term goal but we don’t have the funding or the time to do it right now.

Any helpful suggestions are much appreciated.


Hello Ricky,
as far as I know, the old Enterprise ALE Adapter does not use SAP JCo for RFC communication, but a very old C++ IDoc library. SAP stopped support for that old IDoc library back in 2001. And of course this library is not Unicode-enabled.

So most probably, when this library receives IDoc data in Unicode format, it crashes, because all fields are twice as long as expected, leading to memory buffer overflow.

You may be able to still use that Adapter for receiving IDocs from SAP, if you customize the corresponding RFC destination in SM59 to use “Non-Unicode”.
However, this will only allow ISO-Latin-1 data to be transmitted. (See also SAP Note 793847 for more details about ALE communication between Unicode SAP system and Non-Unicode receiver. There are tricks to also enable other codepages like Kyrillic or Japanese, but this would lead to far here. The SAP Note describes a few options, if you need other codepages.)

Cheers, Lanzelot