On Wednesday, 03/18/2015 at 06:04 EDT, Rob van der Heij <rvdh...@gmail.com> wrote: > Sorry to complicate it, but that check is not conclusive. When Linux does > not run ntpd, the time on Linux will match the time on z/VM (give or take > time zones). But when they still match the wall clock time a few weeks > after the IPL of z/VM, you know STP is steering the clock.
Someone said to me that they thought zLinux will still use jiffies in certain cases, which can cause "now" to have some significant variation. Perhaps Those Who Know will speak up. I've done more thinking, studying, and talking to Development about this, and I now believe that TOD clock steering does NOT deal with leap seconds. This is because the TOD clock is steered to Coordinated Server Time (CST), not UTC, and CST is computed by adding the leap seconds to UTC. UTC can be determined by manual input or from NTP. So if you set leap seconds to zero, you have UTC but only until the next leap second. But Development is inquiring of PE to verify. Architecturally, the z TOD clock is ahead of UTC by the number of leap seconds contained in the UTC clock. And please know, too, that it is the STP feature that steers the TOD clock to the external time source. Without it, you have to re-IPL z/VM to get the LPAR TOD clock resync'd with the physical TOD clock, which is adjusted to match the time obtained by the SE from the external time source at model-dependent intervals. Leap seconds become important when you start reaching back in the time. If you reach past the most recent leap second insertion point, the wall clock or TOD clock conversations start being off by one second per insertion. For many things, that's close enough. For others (e.g. financial institutions) the time standards are established by regulatory agencies. But you must be careful. If you think about this too hard, you will create a tear in the time-space continuum and fall in. Just remember that "The music is reversible, but time is not. Kcab nrut, kcab nrut, ...." Alan Altmark Senior Managing z/VM and Linux Consultant Lab Services System z Delivery Practice IBM Systems & Technology Group ibm.com/systems/services/labservices office: 607.429.3323 mobile; 607.321.7556 alan_altm...@us.ibm.com IBM Endicott ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 ---------------------------------------------------------------------- For more information on Linux on System z, visit http://wiki.linuxvm.org/
