1 |
On 01/18/2018 08:13 AM, Alec Warner wrote: |
2 |
> On Thu, Jan 18, 2018 at 6:46 AM, Anthony G. Basile <blueness@g.o |
3 |
> <mailto:blueness@g.o>> wrote: |
4 |
> |
5 |
> Hi everyone, |
6 |
> |
7 |
> I'm trying to design an update system for many identical Gentoo systems. |
8 |
> Using a binhost is obvious, but there are still problems with this |
9 |
> approach. |
10 |
> |
11 |
> Unless there's some magic I don't know about (and this is why I'm |
12 |
> sending this email) each machine still needs to have the portage tree |
13 |
> installed locally (1.5 GB) or somehow mounted by a network filesystem |
14 |
> (which is not practical if the machines are not on a local network). |
15 |
> |
16 |
> |
17 |
> +Zac |
18 |
> |
19 |
> I don't believe this is true; with the correct binhost configuration |
20 |
> portage should: |
21 |
> |
22 |
> 1) Contact the binhost to get a dump of packages available on it. |
23 |
> 2) Use that dump to create a 'virtual portage tree' (bindb). |
24 |
> 3) Use the bindb for package discovery (is foo available) and dependency |
25 |
> calculation. |
26 |
|
27 |
It's already possible if you use PORTAGE_BINHOST and create a dummy |
28 |
profile to satisfy portage. I've filed this bug to create a convenient |
29 |
option for this: |
30 |
|
31 |
https://bugs.gentoo.org/644990 |
32 |
|
33 |
> |
34 |
> Furthermore, each machine would have to run emerge locally to do the |
35 |
> calculation of what packages need updating. |
36 |
> |
37 |
> |
38 |
> The good news is that with a bindb; the package tree itself is much |
39 |
> smaller. ::gentoo is 23000 packages. |
40 |
> But you bindb is only as large as you build binpkgs for and publish |
41 |
> them. So say stage3 + some other stuff. |
42 |
> Around 500 packages perhaps. |
43 |
|
44 |
Doing the calculation on the client side is fine, since binary package |
45 |
calculations are much simpler than source-based ebuild calculations. |
46 |
-- |
47 |
Thanks, |
48 |
Zac |