We have had a pmr opened with TWS support for the past 45 days and they were able to reproduce the problem we reported but they do not feel they need to open an APAR for this because fixing the problem is beyond the current design of the product and therefore not APARable. We have continued to ask for this apar anyways and their latest attempt at appeasing us consists of a documentation apar updating an obscure 'installation considerations' paragraph which will eventually be changed to state that deleting a data set may adversely affect a job being restarted by TWS. I'm beginning to feel that by continuing to persist that I am being unreasonable.
-----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of Brian Kennelly Sent: Wednesday, September 08, 2010 5:50 PM To: [email protected] Subject: Re: S737 Abends, Deleted GDG Data Sets and TWS On Wed, Sep 8, 2010 at 15:53, Tabor, Rich <[email protected]>wrote: > The short answer I think is yes - yes it is TWS's fault. Yes, as > suggested, if I specify the absolute GDG data set name in the JCL, I > see only the enqueue for the absolute GDG data set and no enqueue for > the GDG base. And MVS/DFP does not allow the data set to be deleted > by another tso user or batch job in this instance (as we all hope and > expect). I don't know for sure if it has to do with enqueuing on both > the GDG base and the actual gdg data set. I don't know why it is > possible in some circumstances for a TSO user or some other job to > delete a gdg allocated to a job restarted by TWS, but TWS is > apf-authorized and is modifying JES control blocks so I can't see > another culprit here. I'm just making sure that the problem is on > record so that someone running into this in the future will find that it has > happened before. It is very likely a basic error in the TWS restart. I remember another vendor's restart product that had a similar problem. it would reset the SWA blocks for a GDS to reflect the absolute generation, as John described, but it did so after MVS had built its enqueue table, so the GDS did not get enqueued. The programmer had to add code to acquire the enqueue and update the enqueue table. When I wrote my restart code, I made sure to avoid that error. I adjust the base, but not the relative generations on generation datasets. ---------------------------------------------------------------------- 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 ---------------------------------------------------------------------- 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
