> -----Original Message----- > From: IBM Mainframe Discussion List On Behalf Of Brian Peterson > > I suggest opening a PMR with IBM regarding this issue.
Might just do that when current projects are completed. However, after I APPLY'd the first batch of PTFs, a subsequent RECEIVE ORDER for RECOMMENDED retrieved all remaining RSU* PTFs, so it seems the problem occurs only with a "virgin" CSI. Seems to suggest that the GIMXSID process failed to build a (correct) bitmap of installed service, since there wasn't any service installed yet. > FWIW - my guess - a side effect of a problem similar to and fixed by > IO10260. In my opinion, this guess has a 1/3 chance of being right, and a > 2/3 chance of being wrong. > > Your problem illustrates for me why IBM should not have ever implemented > selection criteria in RECEIVE ORDER. Every RECEIVE ORDER job I have ever > run has specified CONTENT(ALL). > > My reasoning.... > > 1) You're going to eventually get each PTF. Why wait? Not necessarily true: "Institutional paranoia", for want of a better term. For example, the continuing saga of APAR -> PTF -> PE for Catalog Auto-tune is not exactly a confidence-builder. :-) There's still a noticeable aversion to applying any PTFs without which the software simply will not run. > 2) Disk space is cheap. What if you need a PTF that came out three weeks > ago "right now" to resolve a problem, and for whatever reason, internet > download is broken or slow? > > 3) The attribute that helps you manage your PTFs is NOT whether or not > you've downloaded the PTF. Rather, the attribute that helps you manage your > PTFs is the SOURCEID assigned. If you like to age PTFs, watch the SOURCEIDs > - that's the aging process today. Good points, to be sure.... > 4) The added complexity of the CONTENT filter command for RECEIVE ORDER > inevitably results in service instability. Each additional feature requires > either additional testing, or runs the risk of error due to lack of testing. > Useless features make the overall product less stable. I'm not yet able to conclude the feature is "useless", though it certainly can be misused. -jc- ---------------------------------------------------------------------- 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
