Re: getdate

[zdrysdal@diagnostic.co.nz]

HI

I did as you suggested and changed /etc/rcS.d/S50hwclock.sh to
K50hwclock.sh and once rebooted the time was 20:00 instead of 0800... hence
it didn't work.   One of the 3 servers doesn't even have the S50hwclock.sh
file.





debian-user@lists.debian.org on 11/10/99 13:50:18

To:   Zane Drysdale/Diagnostic labs/64, debian-user@lists.debian.org
cc:
Subject:  Re: getdate




zdrysdal@diagnostic.co.nz (zdrysdal@diagnostic.co.nz) wrote:

> where can i find the app getdate?  I have 3 debian fax servers which, via
> cron, reboot twice a week.  When they come back up they are 1 hour behind
> and i have to manually reset the date.  getdate may allow me to kinda fix
> this problem.

I've found a problem(*) with Debian's hwclock.sh scripts.

jekyll:~$ /bin/ls -1 /etc/rc*.d/*hwclock*
/etc/rc0.d/K25hwclock.sh
/etc/rc6.d/K25hwclock.sh
/etc/rcS.d/S50hwclock.sh

Mine are modified.  The hwclock.sh links in rc0.d and rc6.d are called
when you're shutting down or rebooting.  The ones that Debian put there
start with an "S" (S25hwclock.sh), which means they're called with
a "start" parameter when shutting down or rebooting.  This does the
opposite of what you want it to do -- it reads from the hardware clock.
By renaming them to start with "K" (as above), I made sure they *write
to* the hardware clock when I'm shutting down.

This may not fix all of your problems, but it's a good start.

(*)Hey, where's my bug?  I sent it in... I know I did, because I got
a copy.  But it doesn't appear to have been received.  Hmm... I'll
resend it.


Jean-Yves BARBIER (jybarbier@wanadoo.fr) wrote:

> You could fix it by installing package xntp3 on a server, but be
carefull, > if your time server is down you'll have troubles to starts the
fax machines.

If this is likely to be a problem, perhaps you should use the ntp package
from potato instead.  Either that, or comment out any calls to ntpdate in
/etc/init.d/xntp3.  The (x)ntpd daemon itself won't cause any problems --
it will run patiently in the background, even if the time servers are
not reachable.

--
Greg Wooledge                    | Distributed.NET
http://www.distributed.net/
wooledge@kellnet.com             | because a CPU is a terrible thing to
waste.
http://www.kellnet.com/wooledge/ |

(uuencoded file att1.unk follows)

begin 644 att1.unk
M+2TM+2U"14=)3B!01U`@4TE'3D%455)%+2TM+2T-"E9E<G-I;VXZ($=N=5!'
M('8Q+C`N,"`H1TY5+TQI;G5X*0T*0V]M;65N=#H@1F]R(&EN9F\@<V5E(&AT
M='`Z+R]W=W<N9VYU<&<N;W)G#0H-"FE%645!4D5"04%91D%J9T).1#1!0V=K
M46%B5FYU-E0K0FI0;$UW0V5*;T@Q44)31VIF2EDP<%5E=V)E,V)O;W<-"F93
M;T%O3BMU;CA725(O9#A31FQ.55)S*TU.,$5.=S1$#0H],V=W+PT*+2TM+2U%
93D0@4$=0(%-)1TY!5%5212TM+2TM#0H-"@``
`
end