On Tue, 2013-01-29 at 11:23 -0700, Al Stone wrote:
> On 01/29/2013 11:09 AM, Mark Salter wrote:
>> On Tue, 2013-01-29 at 14:12 +0800, Chen Baozi wrote:
>>> On Tue, Jan 29, 2013 at 12:52:24AM -0500, Mark Salter wrote:
>>>> On Tue, 2013-01-29 at 09:05 +0800, Chen Baozi wrote:
>>>>> Thanks, Mark. I've switched to another machine which has better
network
>>>>> connection. And It finally works, :) I guess something I've
checkout before
>>>>> was corrupt due to both my poor network connection and big size of
the
>>>>> rootfs.
>>>>>
>>>>> BTW, seems it is not up-to-date? I mean there is no /stage3 under the
rootfs.
>>>>
>>>> Hmm. I looked closer at the tree I checked out and it too had a head at
>>>> commit 3e7ab1bee31082a0 from Al on Dec 31. I'm at a loss to explain
what
>>> It is the same with my local git tree.
>>>
>>>> is going wrong. I have a local tree that appears to be up to date with
>>>> the repo on
fedoraproject.org but the log shows 18 commits which
aren't
>>>> in the tree I cloned from
fedoraproject.org repo earlier today. I'll
try
>>>> and see if I can sort it all out tomorrow.
>>> Thanks a lot.
>>>
>>> Baozi
>>
>> Al, this may be something for you to look at. If I clone the repo using
>> http, I get 3e7ab1bee31082a as the lastest commit. If I use ssh+git, I
>> get everything correctly.
>>
>> --Mark
>>
>>
>
> The only thing I can think of right now is that 'git update-server-info'
> may not be getting run on updates to the repo for some reason. I've
> run it by hand to see if this changes the behavior -- let me know if
> you now get the most recent commit. If not, I'll have to poke the
> infrastructure folks. The short term workaround is to use ssh+git,
> as you indicated.
>
> Infrastructure and I have been struggling with the underlying problem
> that http and git and SELinux are in a bit of a disagreement on how
> access should work -- and it's http that's been losing out so far.
Yay. A clone using http did get the latest bits.
Well, we now know the workaround; do the commit, then ssh into fp.o and
run git update-server-info in the repo. Ugly, but it works.
I'm not sure this is fixable, but I'll poke the infrastructure folks
and see what we can do.
--
ciao,
al
-----------------------------------
Al Stone
Software Engineer
Red Hat, Inc.
ahs3(a)redhat.com
-----------------------------------