1 |
On Fri, 2004-08-06 at 14:06, Joshua J. Berry wrote: |
2 |
> - Automated setup and install (a la Solaris' Jumpstart). This involves several |
3 |
> aspects, such as the creation of a custom profile, precompiling binaries, |
4 |
> automatic partitioning/formatting, automatic configuration, tying into |
5 |
> Catalyst to create an appropriate boot CD to run the install, ... |
6 |
|
7 |
In most cases, a profile is not needed. The profile only specifies the |
8 |
"system" portion and default virtuals. While I can see how some would |
9 |
want it, I don't think it would be a requirement. The pre-compiled |
10 |
binaries are a must. I could see this working in many ways, all of |
11 |
which would need to be discussed. A proper "rollout" tool would not |
12 |
require a CD, so there would be no need to involve catalyst for such a |
13 |
task, though there's no reason to not allow it. I would instead see |
14 |
catalyst taking the role of creating the binaries. Installing new |
15 |
machines would be carried out using industry standard methods, such as |
16 |
PXE to boot an image that downloads from TFTP and loads up a NFS root |
17 |
with an installer image. The configuration would be decided before |
18 |
install by the script, with possibly a final "setup" step after first |
19 |
boot to complete a very small number of configuration options. |
20 |
|
21 |
> - Configuration management across many machines ... how can one easily handle |
22 |
> changes to files in /etc, and propagate them to clients? |
23 |
|
24 |
How does one do this in Red Hat? SuSE? |
25 |
|
26 |
Currently, there just isn't anybody really doing this well. Having this |
27 |
capability would definitely position Gentoo as a leader in the area of |
28 |
workstation and server management. |
29 |
|
30 |
> - Allow users to create and maintain a custom Portage tree (derived from our own |
31 |
> base tree) which they can use to administer their other machines. |
32 |
|
33 |
This is too simple. We already provide everything a user would want to |
34 |
be able to do this. Perhaps I am missing something. Could you |
35 |
enlighten us a bit more on this? |
36 |
|
37 |
> - Probably something else I forgot. ;) |
38 |
|
39 |
A stable portage tree for each release would be a requirement. |
40 |
|
41 |
> While most of these are possible with shell scripts and the like, there are no |
42 |
> official Gentoo tools to do this. I think it would be great for Gentoo to |
43 |
> create some, because I think this is one area in which we can really shine. |
44 |
|
45 |
I agree. At the same time, I realize that Gentoo is still very young, |
46 |
and we have a long way to go and a lot of growing up to do. The main |
47 |
thing I notice about a lot of people is they want to take Gentoo and to |
48 |
turn it into some huge enterprise solution, but they don't want to |
49 |
invest the money nor manpower to do so. Red Hat has capital. They have |
50 |
employees. Every single one of us are volunteers, and honestly, until |
51 |
that changes, there is no way we will ever be able to meet the needs of |
52 |
the enterprise, simply because corporate customers will want things from |
53 |
us that are not fun nor interesting, which means they will not be things |
54 |
we will want to do. |
55 |
|
56 |
I believe that to achieve these goals, we have to start by taking baby |
57 |
steps. Perhaps we should bring up some things at the next manager's |
58 |
meeting? To be honest, it looks like the first steps are going to be |
59 |
the easiest, and at the same time the hardest. We have to decide where |
60 |
we want to go and lay out a plan. The hardest part is we have to stick |
61 |
to that plan and not falter. |
62 |
|
63 |
> I should re-emphasize that there is quite a lot of demand for this, especially |
64 |
> in mid to larger companies which have hundreds or thousands of machines that |
65 |
> they'd like to put Gentoo on, but can't because the setup process is so |
66 |
> involved. I don't remember exactly how many people asked me about it at |
67 |
> LinuxWorld, but enough did to make me remember. :) |
68 |
|
69 |
There was a lot of demand for it at LWCE NYC, yet nothing really came |
70 |
about from it. I encourage anyone interested in the possibility of |
71 |
growing up from a "hacker" distribution and into something the corporate |
72 |
world respects to try to work together on this and make something |
73 |
happen. |
74 |
|
75 |
> So ... what are your thoughts on this? |
76 |
|
77 |
We need to move on it, and not talk about it, then let the topic die off |
78 |
and forget about it as has been going on for months now. |
79 |
|
80 |
> On a related note, avenj did mention the Installer project to me, and gave me a |
81 |
> link to their CVS, but it hasn't been touched in a few months. Anyone know what |
82 |
> their status is, and what specifically they're trying to accomplish? |
83 |
|
84 |
I believe their charter was to create a nice installer that could be |
85 |
used to install Gentoo in various fashion, using GRP as a base and |
86 |
allowing for rapid deployment of servers and workstations based on |
87 |
pre-defined criteria. |
88 |
|
89 |
-- |
90 |
Chris Gianelloni |
91 |
Release Engineering QA Manager/Games Developer |
92 |
Gentoo Linux |
93 |
|
94 |
Is your power animal a penguin? |