>Since you're referring to problems with the WLM policy, I guess you had some classification rule catching the initial job but found that the child is not caught by this rule.
That was the exact problem. Because any other sub-system would switch into USS, the rules for that sub-system would come into effect. The OMVS service class was similar to TSO, so we would end up with test batch running highere than production, or onlines/stc running lower than where they should be, because they switched into USS. HOD, TN3270, and TCP/IP all suffered, until we were able to diagnose them, using what little tools were available back then. >The FTP daemon comes to mind: It forks once or twice at startup and the >parent(s) terminate there after leaving only the final child in place. That was also one of the problems. >But note that the ftpd code itself let's the parents terminate. That's not what I was told at the time. The VTAM sysprog (who was also just learning about it, as well) told me what he thought it was due to. And, that misunderstanding is what I stated in a previous post. It was 1997-1999, and we were all learning what it was all about. This was before the major re-write of TCP/IP, and even IBM seemed to be learning how to deal with UNIX System Services. We had a lot of mis-information, back then. - Too busy driving to stop for gas! ---------------------------------------------------------------------- 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
