On Fri, Jun 3, 2011 at 02:10, Robert Haas <robertmhaas@gmail.com> wrote:
> On Tue, May 31, 2011 at 6:40 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>> git.postgresql.org still isn't showing a commit I pushed to master at
>> around 21:55 UTC (now about 45 minutes ago). This is the first time
>> I've ever seen it be more than a minute or two behind, so I'm thinking
>> something is busted. In particular, I wonder if this is relevant:
>>
>> $ git push
>> Counting objects: 105, done.
>> Compressing objects: 100% (55/55), done.
>> Writing objects: 100% (55/55), 4.79 KiB, done.
>> Total 55 (delta 51), reused 0 (delta 0)
>> Auto packing the repository for optimum performance. <------
>> To ssh://git@gitmaster.postgresql.org/postgresql.git
>> fccef77..08779dc REL8_2_STABLE -> REL8_2_STABLE
>> 1d6dd87..ca76a39 REL8_3_STABLE -> REL8_3_STABLE
>> 0699d05..8057b75 REL8_4_STABLE -> REL8_4_STABLE
>> e73bd1e..b26d8fd REL9_0_STABLE -> REL9_0_STABLE
>> ba4cacf..6923d69 master -> master
>> $
>>
>> I don't recall seeing that before, either, and am thinking maybe it
>> confused a trigger or something?
>
> It appears to be caught up now. Probably the master got repacked at
> the same time, and that made it run slower than normal.
Ugh, my bad. I responded to Tom and missed the list :(
The problem was that there were no local refs left after the pack, and
the push script only looked at local refs to figure out when to push
it over. It would've fixed itself at the next commit anybody made, but
I also fixed the script so it looks at packed refs as well, and thus
got it sent over.
--
Magnus Hagander
Me: http://www.hagander.net/
Work: http://www.redpill-linpro.com/