F20 Self Contained Change: Remove deprecated calls of using ntpdate in favor of ntpd

[Paul Wouters]

On Wed, 17 Jul 2013, Chris Adams wrote:

> Once upon a time, Paul Wouters <pwouters at redhat.com> said:
>> That's easiest said then done. It takes a lot of queries before you hit
>> pool.ntp.org. And then you have to 1) ensure no one else uses those DNS
>> answers and 2) flush the cache when enabling DNSSEC.
>
> Well, it would be an ugly hack, but you could use something like "dig
> +short +cd pool.ntp.org" to get some IPs (+cd disables validation for
> this request only).

I understand the query. But you would either need to bypass the local
dns caching resolver or flush the cache afterwards. The second option has
a race condition, but the first has the problem that we are trying to reduce the
number of applications that modify /etc/resolv.conf to one (NM).

I'd rather not make "dig" a dependancy, but use libunbound directly with
a CD flag.

>> That's why for a simple "reboot", we could save the time to have some
>> approximation of time when we start (if we have no realtime clock or
>> see the time is 1970 of 2000)
>
> If the root filesystem is ext4, you could use the "last write time" as a
> starting point.

That's not very compatible with other fs'es. What if someone is
upgrading from ext3? Or using brtfs? Or something new? I'd rather see a
more generic method of writing a timestamp to a well known location.

Beyond the saved timestamp, I think I have a preference of only using
DNS queries to resolve this, making it a self-contained issue.

Paul