ntpd flails on machine with large clock error...

["Joe Emenaker" <jemenake@genesiscom.com>]

I've got a machine that's 12800 seconds behind the rest of the world. I'd
like to fix that.

With normal execution of the daemon, it exits immediately after writing to
the logs something to the effect of "clock over 1000s off... set clock
manually...".

However, in this case, I'm more in need of a *smooth* transition to the
proper time than I am in need of a *prompt* one. I don't mind if ntpd needs
a day or two to bring things into line...

So, I tried the "-g" option, which is supposed to disable the "1000s sanity
check". This causes ntpd to run for about 5 minutes before exiting with the
same log message as before. In the intervening 5 minutes, no clock
adjustment happens.

Does anybody have any ideas?

- Joe