On Mon, 2012-01-23 at 16:07 -0500, Paul Gilmartin wrote: > And the TOD just drifts?
Yeppers. It's not remotely as bad as it was in (say) the pre-CMOS days. The clock doesn't move far enough from wall clock to be very noticeable. > How does this affect critical timestamps > kept in GMT? What do z/OS UNIX utilities (FTP?) show? GMT will drift with the seasons of course, but... > May I assume you don't have a sysplex which requires extremely > precise synching of the TODs? You presume correctly. We're just a tiny shop, what a pal of mine used to call "family data processing". Heck, I'm still running monoplex (though those pesky ZFS and PDSE datasets will eventually force me to do something about that). > Or with the system quiesced, do you actually issue an SCK instruction. > (It's not Shmuel's dog.?) Ohno. But it was only fairly recently that IDMS learned about CVTLDTO. IDMS's idea of local time varied from z/OS's idea. That was a small nuisance. > Is the hardware used by STP well enough documented in the PoOp > (PrOp?) that one could write RYO (E)TOD steering code, driven by NTP? I thought that STP was LIC, and contained some enabling bits? -- David Andrews A. Duda & Sons, Inc. [email protected] ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
