| 1 |
After a somewhat lengthy discussion among a couple of the infra dev's |
| 2 |
and Nick, we came to some conclusions |
| 3 |
|
| 4 |
(1) A single point of distribution is just a bad idea imho. Imagine a |
| 5 |
critical patch that MUST GET OUT IMMEDIATELY (differing opinions of |
| 6 |
'must get out now' apply). You get thousands, even tens of thousands of |
| 7 |
people hitting patches.g.o and you can wave goodbye to that machine. |
| 8 |
|
| 9 |
(2) The distribution network is there for a reason. Let me be the first |
| 10 |
one to admit that it isnt perfect. Once you place something on |
| 11 |
/space/distfiles-local, it hits two machines before it hits the world, |
| 12 |
taking between an hour and four hours (theoretically) to go public. |
| 13 |
Latency is involved in whichever solution we can propose, short of |
| 14 |
giving everyone write access to our master mirror. |
| 15 |
|
| 16 |
(3) Whether we use distfiles mirrors, or the rsync rotation to |
| 17 |
distribute patches, this has not been decided upon. The idea was |
| 18 |
broached for a couple minutes earlier today but not even a preliminary |
| 19 |
solution was conjured. |
| 20 |
|
| 21 |
Infra has complete control over the rsync.g.o rotation, so coming up |
| 22 |
with ways to improve it that way can be worked out. I am sure this will |
| 23 |
be a topic during the next dev meeting, if not listed, I will definitely |
| 24 |
mention it. |
| 25 |
|
| 26 |
My 2 infra cents...Get irate, berate, celebrate. |
| 27 |
-Jeffrey |
| 28 |
|
| 29 |
On Tue, 2004-10-12 at 16:10, Nicholas Jones wrote: |
| 30 |
> And rewritten for clarity: |
| 31 |
> |
| 32 |
> 1. developer creates an ebuild which needs a patch |
| 33 |
> 2. SRC_URI="mirror://gentoo/patchname http://patches.gentoo.org/patchname" |
| 34 |
> 3. Uploads files to dev.gentoo.org:/space/distfiles-local |
| 35 |
> 4. developer commits ebuild(s) to cvs |
| 36 |
> |
| 37 |
> Infra notes: |
| 38 |
> |
| 39 |
> Immediate availability via the "secondary" host while the primary |
| 40 |
> host has yet to receive the files. Once the file is present on the |
| 41 |
> primary mirrors it may be deleted from the secondary. |
| 42 |
> |
| 43 |
> Ensuring that the patch host is not the primary mirror will be a |
| 44 |
> concern, but a script can be devised to ensure duplication of |
| 45 |
> the patchname and that a primary mirror is listed prior to it. |
| 46 |
> |
| 47 |
> The host for the files should not be accessable for any reason |
| 48 |
> except direct filename downloads. No listings (to discourage |
| 49 |
> setting it as a mirroring source). |
| 50 |
> |
| 51 |
-- |
| 52 |
|
| 53 |
|
| 54 |
-------------------- |
| 55 |
Jeffrey Forman |
| 56 |
Gentoo Infrastructure |
| 57 |
Gentoo Release Engin. |
| 58 |
jforman@g.o |
| 59 |
-------------------- |
Archives