SMFDUMP is used on some of the LPARs at the consolidated site where I work. The program is run via automated job scheduler; frequency varies by how busy that LPAR is.
SMFDUMP has a problem sometimes where it hangs. I haven't seen an enqueue cause. I haven't taken a dump of it yet but I suspect that it is when the SMF datasets on the sandbox system, rarely dumped, are full, and that somehow causes a problem. I strongly recommend the use of the IEFU29 exit. It just needs, at minimum, to issue something like 'S SMFDUMP,DS=' and fill in the SMF dataset name (which is passed to the exit). The LPARS which use IEFU29 have an SMF dump proc which creates the output as a new entry in a GDG, sized to hold the data from one SMF dataset (the largest, if they are not all the same). At midnight automation issues the SWITCH SMF command (aka 'I SMF') to force a switch, which in turn kicks IEFU29. On a daily basis, the GDG entries are gathered into a daily file and the GDG cleared. That job, SMFDAY, is scheduled via an automatic job scheduler. Other daily processing of SMF data is done by jobs which are successors to the SMFDAY job. The IEFU29 / proc method has so many advantages that we will be switching everyone over to it. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
