We were early adopters of SMF logger in order to prepare for a winter SHARE session after a 1.9 ESP. At that time we used it only in the sandbox sysplex where SMF data was throwaway. It's now running also in the development sysplex where SMF data is actually cherished and nurtured. With only a hiccup or two, the process has been reliable if a bit kludgy. The main problems I recall have involved a shortage of DASD space for the offload data sets. Normally there's plenty here, but some (dimly recollected) condition caused an offload failure. I don't know whether the weakness is in SMF's use of logger or in logger itself, but once a 'full' condition occurs and the structure gets disconnected, I've had a very hard time jump starting the process without losing data.
There are rumblings about improvements on the way that should upgrade the prime time status from 'not ready' to 'get set'. Maybe even 'go'. . . JO.Skip Robinson Southern California Edison Company Electric Dragon Team Paddler SHARE MVS Program Co-Manager 626-302-7535 Office 323-715-0595 Mobile [email protected] Mark Zelden <mark.zel...@zuri CHNA.COM> To Sent by: IBM [email protected] Mainframe cc Discussion List <[email protected] Subject .edu> Re: SMF LOGGER - Not Ready for Prime Time 05/11/2009 06:24 AM Please respond to IBM Mainframe Discussion List <[email protected] .edu> On Sun, 10 May 2009 10:47:36 -0500, Jim Marshall <[email protected]> wrote: > But I think IBM >was remiss in not understanding the full ramifications of how they have sites >implement the LOGGER. If they would have talked to some savy users, then >could have made the journey much smoother. That was apparent from the first time real presentation I saw at SHARE. And we've already had plenty of past discussions about that. At this point I still have no plans to implement it since the old way can keep up at our shop and have never had problems with lost SMF data. I do like the concept of multiple logstreams and the convenience it may eventually bring, but for now we'll just split off the different files during our nightly processing. I do have it running in some sandbox LPARs where we don't care about SMF data. In the past, those LPARs just dumped to a GDG in case we wanted the data later or dumped to "DD DUMMY". Mark ---------------------------------------------------------------------- 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
