John, I think you perhaps understood the SMB recommendation backwards. What I think we're suggesting -- well, I'm suggesting -- is to place the dataset on a z/OS SMB share and let the Windows servers fetch it as they wish, when they wish. Schedule deletion from the share every month, week, or whatever, but otherwise don't do much. "The information is available -- come get it from Drive M." This would also probably vividly demonstrate how often Windows servers fail to those who run them, but that's a side benefit. :-)
It's called "pull" instead of "push" (from the Windows server perspective). If they'd like "push," ask for funding -- you need a pusher. PM4Data happens to be my favorite, and it's a lot less expensive than professional time. (Clients are free AFAIK.) All that said, here's something to keep in mind: every time I see "FTP" (or "file transfer" generally) in somebody's architecture I ask about it. It's often a sign of bad architecture. (Not always, but often.) One thing that's true for sure: if there's an online system on one end of the file transfer you've just made it batch. If you transfer a file from an online system the information is out-of-date the moment you do so. So that's the basic question I ask: why isn't the information available online, real-time? And that's my question here: why can't the Windows servers get online, real-time access to these data? Why are you copying files? There could be a perfectly reasonable explanation, but I'm just wondering. FTP should not be mistaken for integration: it's only file transfer. There's also the not so minor issue of information security -- copying simply multiplies your privacy protection challenge. These days Windows servers like to have Web services or ODBC for online access, depending on the source. Mainframes do both very, very well. - - - - - Timothy F. Sipples Consulting Enterprise Software Architect, z9/zSeries IBM Japan, Ltd. E-Mail: [EMAIL PROTECTED] ---------------------------------------------------------------------- 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
