Peter Hunkeler wrote: >>>I am trying to troubleshoot a user's job which is using ADRDSSU. >>That job is DELETING the datasets on the selected volsers. Is that what the >>user wants?
>Does it? The JCL is referring to the tape by VOL=SER=, not via catalog. In that OP's job, I see UNIT=3390, that is DASD, not tape. I agree that job is looking at VTOC, not Catalog. >Non-SMS DASD data sets are deleted from the volume but not removed from the >catalog when located via VOL=SER=. True. I don't like this way of deletion, but is very useful in case of emergency or mismatch ... >I'm not sure what the system does in the case of tape data sets. The data set >cannot be physically deleted from the tape, because there might be more data >sets after the one to be deleted. I never cared to investigate, but I assume >that deleting a tape data set will only remove the catalog entry and leave >alone the physical data on the tape. Yes, I agree with all you say. It is a real PITA to do that deletion on such tapes and their contents. You need to do a tape scan to check the tape and then instead of deletion, you rather copy the datasets you want to preserve to another tape while adjusting the LABEL one by one. Of course, with today's tape management systems with 'stacking' features, no such scenario is needed. >I'm not a storage guy, and I don't have experience with tape catalog systems >(RMM, CA-1, etc). Would those recognize the reference by VOL=SER= and delete >the entry from their inventory? For Control-T, I think the answer is no, but we have setup Control-T that only cataloged and RACF protected tape datasets are used. So I can't test out your scenario. (But there is a Control-T option + RACF profiles which allow/reject such mounts via VOL=SER= statement.) >Curious to know. Damn! Now I am also curious!!! Groete / Greetings Elardus Engelbrecht ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
