My apologies... for testing my ooRexx / mod_rexx / ORexxSQL application, I was forgetting to update both the client and server nodes. I was just updating the server node and testing. The client node was yet running ooRexx 3.2 successfully.
DB2 6.1 FP11_WR21302
Rexx/SQL is from rxsql25_odbc_w32.zip
For generating HTML files... "export the site to HTML" mode, a special Nehemiah Project Rexx program bypasses any need of Apache and simply cranks through every webpage for a particular site and writes the resulting HTML to the file system.
I have confirmed the same sort of crash with both ooRexx 4.1.3 and 4.2-RC builds. The two versions crash building different webpages, however they crash identically with the same query attempting to be executed. (The 4.2.0-RC code actually gets just a little bit further in number of pages generated.) The debug output from the Nehemiah Project is as follows:
20140124 15:20:49 *** DATABaseError *** 20140124 15:20:49 PARESE SOURCE: WindowsNT METHOD D:\Distrib\ObjCUR\LDSRexxSQL.cls 20140124 15:20:49 ERROR LINE: 328 20140124 15:20:49 SOURCELINE: return myResultTable~fetch 20140124 15:20:49 EXCEPTION CLASS: The SQLRESULTTABLE class 20140124 15:20:49 EXCEPTION: Fetch failed 20140124 15:20:49 FUNCTION: SQLCA.FUNCTION 20140124 15:20:49 SQLCODE: -954 20140124 15:20:49 SQLERRM: [IBM][CLI Driver][DB2/NT] SQL0954C Not enough storage is available in the application heap to process the statement. SQLSTATE=57011 20140124 15:20:49 SQLSTATE: 57011 20140124 15:20:49 SQLTEXT: SELECT NULLID.PRODTEXT.TEXT FROM NULLID.PAGES,NULLID.PAGEPRODLINK,NULLID.PRODUCTS,NULLID.PRODTEXT WHERE (((((NULLID.PAGES.ID = '00010214' ) AND (NULLID.PAGES.ACTIVEFLAG = '1' ) ) AND (NULLID.PAGES.ID = NULLID.PAGEPRODLINK.PAGEID ) ) AND (NULLID.PAGEPRODLINK.PRODID = NULLID.PRODUCTS.PRODID ) ) AND (NULLID.PRODUCTS.PRODID = NULLID.PRODTEXT.PRODID ) ) ORDER BY NULLID.PRODTEXT.SORTORDER 20140124 15:20:49 *** CompileLine - RSPSyntax *** 20140124 15:20:49 PARESE SOURCE: WindowsNT METHOD N:\Websites\CGS\www\NehemiahUtil.cls 20140124 15:20:49 ERROR LINE: 327 20140124 15:20:49 SOURCELINE: interpret 'OUTline=OUTline || BP ||' RSP 20140124 15:20:49 RC: 91 ERRORTEXT: No result object 20140124 15:20:49 RSP LINE: self~GetAllProdText() 20140124 15:20:49 INLINE LINE: <?rexx self~GetAllProdText() ?> 20140124 15:20:49 BP LINE: 20140124 15:20:49 AP LINE: 20140124 15:20:49 OUTLINE LINE: 20140124 15:20:49 ERRORTEXT: No result object 20140124 15:20:49 CONDITION NAME: SYNTAX 20140124 15:20:49 CONDITION INSTRUCTION: SIGNAL 20140124 15:20:49 CONDITION STATUS: OFF 20140124 15:20:49 CONDITION DESC: (Nothing)
I just put ooRexx 3.2 back on the client node, and the Nehemiah Project code is able to build all of the HTML's successfully for the site.
I sincerely apologize for not realizing that I was not testing adequately before this time. I indeed realize that MUCH has changed in the ooRexx code base between v3.2 and v4.1.3. :-(
The rexx.exe process starts out drawing approx 53MB of both "Mem Usage" and "VM Size" memory. VM Size is always slightly greater than Mem Usage. When it crashes, it has reached 56MB for both memory indicators. The RxAPI.EXE process seems to be lazy and not draw any additional memory as the program works.
How could I assist in finding the root cause of the resource leak?