From: Matthew Excell (matt_at_possibilityforge.com)
Date: Thu 29 Jan 2004 - 22:39:56 GMT
On Thu, 2004-01-29 at 14:25, Herbert Poetzl wrote:
> 1) should the 2.4 branch be frozen, and if when?
Feature frozen - probably anytime. We have been using 2.6 kernel where
possible for a month now with no major issues (except on x86_64.)
I personally feel that fixes to major bugs, security issues, etc. and
patches to handle release versions of the 2.4 kernel should continue.
(We have one production server still running 2.4 and an older vserver
with no reason to upgrade it - it works well.)
> 2) is it bad if 2.4 and 2.6 branches diverge?
I don't believe so. There is a precedent for this - look at ipchains,
iptables, netfilter. When major sections and features of the kernel
change with a new major release (2.4->2.6 for example) to allow a
better(?) way of doing things, it only makes sense to move with it. No
reason to make sure a jet airplane can use car tires. . .
> 3) is multitude (2.4/2.6/2.6SE) preferred over
> development speed?
I personally think development speed gained (and thereby better
virtualization / better security) is more important than backporting
> 4) is a 2.4/2.6 migration path important to you?
Not so far as the tools are concerned. As long as the virtual server
directories can be dropped more or less intact into the new
installation, I'll be happy. (I can't see why it wouldn't - I mean the
vserver directories now are basically a regular root filesystem - that
shouldn't change too much, right?)
> 5) what branch 2.4, 2.6 or 2.6SE would you prefer?
2.6 or 2.6SE for our uses.
> 6) what branch/version do you currently use/test?
32 bit 2.6: 1.3.6
64 bit 2.6: 1.3.6 - soon hopefully
32 bit 2.4: 2.4.19 6+ month old vserver (I don't remember which)
64 bit 2.4: testing 2.4.25pre-7 1.3.6
-- Matthew Excell <matt_at_possibilityforge.com> The Possibility Forge, Inc.
_______________________________________________ Vserver mailing list Vserver_at_list.linux-vserver.org http://list.linux-vserver.org/mailman/listinfo/vserver