[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: A few observations about systemd

Hi Jon,

On Mon, Jul 18, 2011 at 10:30:15AM +0100, Jon Dowland wrote:
> Likewise,  a recent kernel does not seem like a problem, and cgroups seems
> like a fairly core part of what systemd does.

There are use cases where requiring the latest kernel would be a problem. 
For example, some virtual hosting environments, such as those using Xen
virtualization, don't give you control over the kernel you're running.  I
seem to remember 2.6.18 being a common kernel vintage in use with Xen, which
is definitely too old to work with systemd; but even if Xen moves forward to
a newer preferred kernel version, systemd could adopt and start to depend on
some other kernel feature down the line and cause the same problem.

The udev+kernel version coupling already gives us maintenance headaches for
distribution backwards-compatibility and upgradeability.  I suppose most
people running Xen can avoid this because it's virtualized and they can get
away without using udev; but when PID 1 won't start, it's a different
proposition entirely.

I guess the same applies for containers like LXC and such - you don't get
your own kernel, you just get your own kernel namespace, and you have to run
PID 1...

So yes, I expect strict kernel requirements from an init system to be a

Steve Langasek                   Give me a lever long enough and a Free OS
Debian Developer                   to set it on, and I can move the world.
Ubuntu Developer                                    http://www.debian.org/
slangasek@ubuntu.com                                     vorlon@debian.org

Attachment: signature.asc
Description: Digital signature

Reply to: