1 |
Thierry Carrez wrote: |
2 |
> Hi folks, |
3 |
> |
4 |
> I would like to get your opinion on Enterprise-oriented desktop |
5 |
> deployment tools for Gentoo Linux (or the lack of). |
6 |
> |
7 |
> As a small company CIO, I deployed Gentoo on a small scale here but |
8 |
> quickly ran into scaling problems and the lack of tools to help. |
9 |
> |
10 |
> There is no obvious way to freeze a Portage tree (or to design a |
11 |
> specific profile) for testing on a golden workstation, to build a set of |
12 |
> update packages (ServicePack) and push it to the workstations, or to |
13 |
> have centralized accountability of what's installed where. There is no |
14 |
> easy way to avoid having to keep a synchronized copy of the portage tree |
15 |
> on all systems, even when using yourown-binaries. |
16 |
|
17 |
Network mountable trees seem to work well enough although an emerge |
18 |
--metadata is still required on clients. |
19 |
I would disagree also on the profile argument. Profiles are very |
20 |
powerful, very details, and have decent manpages as well as literally |
21 |
tons of examples. What specifically is stopping you from rolling your |
22 |
own profile? |
23 |
|
24 |
The rest of that stuff is a generally well known about issue ( at least |
25 |
in the portage community ). Many features of portage-2.1 will be |
26 |
helpful in this type of situation. |
27 |
|
28 |
> With automatic deployments, would we run into difficult-to-solve |
29 |
> etc-update problems ? Should/could the ServicePack system take care of |
30 |
> that ? |
31 |
I wouldn't use etc-update for this on a enterprise rollout personally. |
32 |
If you need config cfengine does a nice job, as well as using |
33 |
cvs/rcs/something-else |
34 |
|
35 |
> |
36 |
> Even in a simpler setup (preprod > production) we don't have the tools |
37 |
> to push a software configuration change from a test machine to a |
38 |
> production one. |
39 |
What exactly are you looking for here? |
40 |
> |
41 |
> What tools are missing ? Is it our job to provide them ? Can it |
42 |
> reasonably be done ? Am I just wrong to want to use Gentoo in that |
43 |
> direction ? |
44 |
Portage needs work; I know the devs are working on it, I know there |
45 |
are other people who are doing there own things. I see a lot of |
46 |
portage-2.1 features that greatly simplify what you are trying to do ( |
47 |
repositories, config rewrite..etc.. ). I think portage and what it |
48 |
covers is a big part of this. Recollecting a conversation with jstubbs |
49 |
about portage he mentioned that he wouldn't want the portage-team to |
50 |
maintain a Enterprise-like distribution program, but that the new API |
51 |
would be great to write one against ;) |
52 |
|
53 |
I also know Chotchki was looking at doing his senior thesis on a network |
54 |
aware portage that did some cool things. A lot of this is just waiting |
55 |
( and helping :) :) ) the portage devs get the work done that needs to |
56 |
get done. |
57 |
|
58 |
I know Cardoe and genstef? are working on a seperate package manager |
59 |
that just handles binaries but uses all the current portage stuff, so |
60 |
you might want to talk to them as well. |
61 |
|
62 |
> |
63 |
> Next week: Gentoo-as-a-metadistribution tools :) |
64 |
> |
65 |
|
66 |
-- |
67 |
gentoo-dev@g.o mailing list |