[Info-vax] Openvms AXP clock problem

[Sandro Altamura]

Dear Stephen,
thanks for your reply. My DEC2K AXP is actually a Jensen machine. I know that it is not a "pure" machine (it can even run WinNT) but when I bought it more than 10 years ago was already coming with OpenVMS installed. I already replaced the 4.5V battery since the old one was exhausted. 
NTPD is not running and the timezone is already set. I switch off the machine only after getting the chevron on the console so the system is definitely shut down.
I guess my next step would be to reinstall vms (maybe to version 7.3) and see if things will go better.




Il giorno venerdì 14 ottobre 2016 23:52:06 UTC+2, Stephen Hoffman ha scritto:

> On 2016-10-14 21:22:16 +0000, Sandro Altamura said:
> 
> > Dear All,
> > I have a strange problem with my DEC 2000 axp running openvms 6.2
> 
> That is the Jensen or related and that's a problematic system to start 
> with, and a very old OpenVMS version.   Get as far forward as that 
> particular Alpha supports — IIRC, Jensen was de-supported somewhere in 
> the V7 range.   That's one of the few Alpha systems that was 
> de-supported by OpenVMS, too.    That box is massively sensitive to the 
> system hardware configuration, and all sorts of odd behavior can arise.
> 
> > The system boots fine and works fine. The problem comes when I shutdown 
> > the system. If I keep it off for 30 minutes and I boot it again the 
> > system time is 30 min delayed. In other words instead of following the 
> > hardware clock the VMS  is continuing with the time of the last 
> > shutdown.
> > I imagine that there is a parameter in sysgen or similar that should 
> > tell the VMS to follow the hardware clock but I could not find it.
> 
> No such parameter exists, nor any such mechanism.
> 
> > I also looked online but the only problems reported where about the 
> > continous request of date and time at boot (this doesn't apply to my 
> > issue)
> 
> Make sure that your timezone is set appropriately or — given the 
> problems with timekeeping that far back — set your system time to UTC 
> and leave it there.   There were some patches to fix bugs in that 
> daylight saving time and related pieces, too.   Half-hour offsets 
> aren't common, but they do exist.   These offsets do not vary by system 
> shutdown time, however.
> 
> If NTP is configured and enabled, then confirm you're not using NTP 
> servers with problematic time values, and confirm that your time 
> synchronization is (not) too far outside of what NTP allows.   That 
> usually doesn't manifest by a half-hour slip secondary to a half-hour 
> power-down, however.
> 
> Assuming you're not powering off OpenVMS hard — performing a hard 
> power-off doesn't update the saved system time, and which can lead to 
> skewed times — and assuming that you're not using an NTP time server, 
> and assuming this isn't secondary to the daylight saving time and 
> timezone settings, then replace your watch battery or Dallas chip or 
> whatever that Jensen box used.
> 
> I'd likely go for a pre-emptive replacement of the battery or the 
> Dallas or whatever that box had, anyway.   The boxes with lithium 
> batteries only lasted five to ten years. then things got weird.
> 
> Related:
> http://labs.hoffmanlabs.com/node/441
> http://labs.hoffmanlabs.com/node/1078
> http://labs.hoffmanlabs.com/node/1895
> 
> 
> -- 
> Pure Personal Opinion | HoffmanLabs LLC