Miklos, >I sent some dumps's to the IBM (with PMR) generated with this options, >and they said , that some storage areas are not in the DUMP
then I should not have picked on you :-) The last time I heard *that* from IBM, I proved to them in excruciating detail that they had no clue what they were talking about. Don't believe everything IBM tells you. First of all, ALLNUC in my opinion should never be specified. For the simple reason that IPCS will always tell you which nucmod you're dealing with. Then you can ask for the maint level of that nucmod. And anyway, the nucmods are only interesting for deep BCP components! In which case having them dumped might expedite pd because IBM doesn't always have to ask what maint level the nucmod is at. Secondly, IBM support in many cases wrongly *assumes* that a dump (a=svcd, which always gets specified, too, despite the fact that it is the default) will show the TCB in question with the PSW from the slip trap. No, not necessarily for an asynchronous dump! And here lies the problem: The customer gets told he did not specify the right options and docs get re-asked for. Useless 'keep-the-customer-busy' strategy. If your dump shows via ip ST command that the slip trap hit at the right module/instruction, then pointedly ask IBM *what* storage is missing. Next ask why *they* did not specify the slip trap so that *that* storage definitely gets dumped! (Possibilities for that: either a=syncsvcd, which will NOT work, if the local lock is held anywhere else in the address space or if local dumping takes too long. Second possibility: explicit definition of the area to be dumped.) In my experience, the excuse of 'some data are not dumped' were always just that: An excuse for stupidity! (Both with regard to 'how to read a dump' and 'how to set a slip trap'.) Thanks for letting me vent! Barbara ---------------------------------------------------------------------- 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
