[Info-vax] satellite one hour ahead

[Stephen Hoffman]

On 2016-10-24 07:15:23 +0000, Phillip Helbig (undress to reply said:

> For the past couple of days, the clock on a satellite has been one hour 
> ahead of the correct time (CEDT) on the other three nodes in the  
> cluster.  If I change it after booting, things are OK until the next  
> reboot, then it is ahead by exactly an hour again.  Any ideas?

OpenVMS timekeeping and timezone support and the daylight saving time 
switch-over has been problematic for many years.    The much vaunted 
idiocy around "compatibility" has utterly screwed the ability to fix 
that mess, too.   But I digress.   Manually reset the TZ, and see if 
that works.   It's common for that setting to be incorrect on ancient 
OpenVMS releases, and V7.3-2 is ~13 years old and this box is almost 
certainly badly under-patched.    With something as problematic as 
OpenVMS timekeeping, it's often simplest and best to switch to UTC.   
If you want to try localtime and lack the patches for the older bits, 
then upgrade to V8.4 — you've been asked to perform that V8.4 upgrade 
before, and for other reasons — and run with something approaching 
current OpenVMS bits, and see if that remediates this.    Or hack the 
TZ database manually as has been discussed before — if there have been 
TZ-related rules changes in your area in the past ~13 years — and see 
if that remediates this.



-- 
Pure Personal Opinion | HoffmanLabs LLC