I'm not familiar with the product. I would suggest that LE tuning is not necessarily your problem. Just for background, which JVM are you running, are you running with shared HFS, and how is your garbage collection? How large is your region? Are you running this from JCL or from OMVS? This sounds very IO intensive. I think there is some data out there regarding the tuning of the XML parser (although it may be WebSphere specific). I'll see what I can find.
Thanks, Aaron On Mon, 23 Oct 2006 10:13:23 -0500, Chase, John <[EMAIL PROTECTED]> wrote: >Hi, All, > >Is anybody here running a product "Whitehill<xml>Transport"? We're >trying to get it to run "efficiently" on z/OS UNIX, but so far the best >performance we've been able to coax out of LE and Java for this >application is "abysmal" (e.g., 7 minutes for a job that "should take" >about 10 seconds). Heretofore my only exposure to Java was as the hot >brown liquid that one drinks to awaken or stay awake. > >I've already tried increasing the storage-related LE parms; that only >made things worse (the 7-minute "benchmark" job's elapsed time INCREASED >to 1:32:57 (that's more than 1½ hours)). > ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
