On Thu, Apr 05, 2001 at 05:31:27PM -0500, Kevin van Haaren wrote: > At 4:28 PM -0700 4/5/01, Jason E. Stewart wrote: > >It would be sooooooooo.... nice if they did. Every time this happens > >to me, it kills postgres, because the negative dates corrupt the > >pg_control file. Luckily I have a spare, uncorrupted one lying around > >for just such an occasion, but man, the first time it happened I was > >freaking out. Also sendmail refuses to work. Since I'm only stupid > >enough to do this about once every three months, I've forgotten that > >it will happen, and I get really irritated because sendmail is > >broken.... and then I remember. > > > >jas. > > what about an init script that runs just after the network is up, but > before postgres, that checks the date and if it's < 2001 force it to > 1/1/2001 then run ntpdate to get the real date?
ntpdate runs way to late in the boot process, i would edit its
initscript to add something like this:
if [ `date +%Y` -lt 2001 ] ; then
date 010100001970
fi
[rest of normal ntpdate script]
this will pull the date up to epoch and then ntpdate should be ok with
updating the time to the correct value, it will require network
connectivity however.
also
update-rc.d -f ntpdate remove
update-rc.d -f ntpdate start 41 S .
> just a work around for a poor design, but the best i can think of.
its a hardware flaw which is hard to really fix... only apple can.
--
Ethan Benson
http://www.alaska.net/~erbenson/
pgp7ZJMMz1ogg.pgp
Description: PGP signature
