1 |
On 01:45 Sat 24 Jan , Robin H. Johnson wrote: |
2 |
> On Sat, Jan 24, 2009 at 01:39:43AM +0000, Mike Auty wrote: |
3 |
> Secondary delay is that I asked for some input on hardware and tuning re |
4 |
> managing a very large repo, and still haven't got a response from my |
5 |
> contact (at kernel.org). Regardless of the tuning, we're going to need |
6 |
> to figure out a box with lots of RAM for the repo size. |
7 |
|
8 |
There was a conversation on #osuosl last week about gitweb setup for |
9 |
performance, which I attached. |
10 |
|
11 |
> > There seem to be two front runners, 1 repo/package and 1 repo/tree. The |
12 |
> > pros and cons are as set out in your file (the flat tree option pros |
13 |
> > don't seem to outweigh the cons, so I haven't really considered that). |
14 |
> |
15 |
> 1 repo/tree is the only one that will fly, see past discussion on the |
16 |
> list. |
17 |
|
18 |
I'm still not convinced of that. The alternative would require a patch |
19 |
to git-submodules, some additional tools support for grabbing additional |
20 |
repos with checkouts and repoman, but it has some compelling advantages |
21 |
(like non-devs don't need 1 GB of web space / bandwidth to post their |
22 |
forks of the tree). |
23 |
|
24 |
On the other hand, it would be the easiest, and perhaps that's reason |
25 |
enough, because the people working on it don't have the time to do |
26 |
anything more. |
27 |
|
28 |
-- |
29 |
Thanks, |
30 |
Donnie |
31 |
|
32 |
Donnie Berkholz |
33 |
Developer, Gentoo Linux |
34 |
Blog: http://dberkholz.wordpress.com |