1 |
On 21 January 2015 at 22:44, Rich Freeman <rich0@g.o> wrote: |
2 |
> On Wed, Jan 21, 2015 at 9:00 AM, Sam Bishop <sam@××××××.email> wrote: |
3 |
>> |
4 |
>> I don't see why it can't be all the combinations, the issue is |
5 |
>> storage, and the storage costs could be a lot lower than expected |
6 |
>> given how hard it is to guess. |
7 |
> |
8 |
> I don't believe that binpkg filenames contain the use flag settings, |
9 |
> and I'm not sure that given multiple copies of a binpkg with different |
10 |
> filenames portage goes through them and figures out which ones are |
11 |
> which. This isn't an area I have looked into seriously. However, it |
12 |
> obviously would be a blocker for getting what you propose to work, |
13 |
> even theoretically. |
14 |
> |
15 |
|
16 |
I'll quote from the binpkg docs: |
17 |
>> Next to these, portage will check if the binary package is built using the same USE flags as expected on the client. If a package is built with a different USE flag combination, portage will either ignore the binary package (and use source-based build) or fail, depending on the options passed on to emerge |
18 |
|
19 |
So I'm fairly sure that implies they can coexist based on the |
20 |
directory structure. - |
21 |
http://wiki.gentoo.org/wiki/Binary_package_guide#The_PKGDIR_layout |
22 |
|
23 |
One big concern would be having a HUGE Packages metadata file and |
24 |
making the look up too slow. I'm not sure how big that file could get |
25 |
before things became an issue. |
26 |
http://wiki.gentoo.org/wiki/Binary_package_guide#Pulling_packages_from_a_binary_package_host |
27 |
|
28 |
> |
29 |
> I don't really see the value in having EVERY combination of use flags |
30 |
> on call though. |
31 |
> |
32 |
> Practically speaking I doubt this could be done. You're talking about |
33 |
> a LOT of combinations. |
34 |
> |
35 |
> However, I think it would be very useful to have a binpkg repository |
36 |
> all the same. Perhaps have one for each of a few common profiles with |
37 |
> the default flags. That alone would be a significant undertaking. |
38 |
> |
39 |
> Just about everybody who has talked about running Gentoo in a |
40 |
> datacenter has set up a binpkg repository. They may very well deviate |
41 |
> from the default USE flags, but for the most part they try to keep |
42 |
> their systems identical. They would build updates as binpkg, install |
43 |
> to a test system, and after testing deploy them to production and that |
44 |
> would of course go quickly. |
45 |
> |
46 |
> I have a script I use to build binpkg nightly for the day's updates. |
47 |
> That lets me review updates and deploy them quickly. Any rebuilds/etc |
48 |
> still take time, but the bulk of my updates are very fast this way |
49 |
> with minimal time spent staring at the screen. This would be another |
50 |
> route to take if your really did need highly varied deployments. |
51 |
> |
52 |
> -- |
53 |
> Rich |
54 |
> |