nanog mailing list archives

Re: Linux: concerns over systemd adoption and Debian's decision to switch


From: Doug Barton <dougb () dougbarton us>
Date: Thu, 23 Oct 2014 16:51:21 -0700

On 10/23/14 4:01 PM, Simon Lyall wrote:
On Wed, 22 Oct 2014, Stephen Satchell wrote:
On 10/22/2014 08:20 PM, Simon Lyall wrote:
On Wed, 22 Oct 2014, Miles Fidelman wrote:
And maybe, you should check out some of the upstream bug reports re.
systemd interactions with NTP.

If you think the current situation is all good then maybe you should
look at other bugs for ntp. eg this one I that affected me with Ubuntu
Disktop. They only run time syncing when the network is bounced so if
you have a stable network then your machine will never sync:

https://bugs.launchpad.net/ubuntu/+source/ntp/+bug/1178933
[..]
I'm a long-time user of NTP, and what you are asking for is a no-good
way of doing things.  What you are supposed to do is use the ntpdate(8)
utility *ONCE* on boot to initially set the system clock, then you have
ntpd(8) running to do two things for you:  sync up to one or more time
sources, and discipline the local clock.
[..]
That's the SERVER way of running a time synchronization.  So it would
appear that you have a quarrel with GUI support, not with NTP itself.

What my point was is that the "simple default for end users" [1] is
already significantly broken in Ubuntu (that is just one bug that bit
me, there are plenty of others).

The systemd system seems to offer and improvement on the existing
"simple default" setup while still enabling experts to run a full ntpd
install if they wish.

[1] - I know how to setup and run ntpd, I didn't expect to need to do it
on my workstation however.

If you are actually arguing that because Ubuntu made a mistake on how the "Internet time synchronization" option is configured, therefore we need systemd, you need to rethink your position. :)

FWIW, the problem you're describing with that option is real, and was revisited in later versions. As of 14.04 it was still broken, but for a different reason having to do with permissions on the ntpd install. However, fixing that problem doesn't require systemd, it requires fixing *that* problem.

I am not against systemd per se, I honestly don't know enough about it to form an intelligent opinion. The line of reasoning (I believe to be) espoused here is quite concerning, "If there is a problem, we need to bring the solution into systemd." To the extent that's accurate, it's overwhelmingly likely to be wrong.

I could say a lot more about Unix system design philosophies from my time in the FreeBSD project, but this thread started off-topic, and has only gotten worse. :)

Doug




Current thread: