Обсуждение: Buildfarm broken for 9.3 and up
My push to branches 9.3 and up seems to have broken the buildfarm with this: error: object file /home/pgfarm/buildroot/pgmirror.git/objects/93/d7706cbf2ce58e63ab8bbc9d16453b2c792ed4 is empty error: unable to find 93d7706cbf2ce58e63ab8bbc9d16453b2c792ed4 fatal: SHA1 COLLISION FOUND WITH 93d7706cbf2ce58e63ab8bbc9d16453b2c792ed4 ! fatal: index-pack failed Your configuration specifies to merge with the ref 'master' from the remote, but no such ref was fetched. I don't see any problems on my machine, although when I attempt to checkout the referenced SHA1 value on branch REL9_3_STABLE, I get this: kgrittn@Kevin-Desktop:~/pg/REL9_3_STABLE$ git checkout 93d7706cbf2ce58e63ab8bbc9d16453b2c792ed4 fatal: reference is not a tree: 93d7706cbf2ce58e63ab8bbc9d16453b2c792ed4 I can't find that SHA1 in my git log. I have no clue what to do about this. -- Kevin Grittner EDB: http://www.enterprisedb.com The Enterprise PostgreSQL Company
Kevin Grittner <kgrittn@ymail.com> wrote: > My push to branches 9.3 and up seems to have broken the buildfarm > with this: > > error: object file /home/pgfarm/buildroot/pgmirror.git/objects/93/d7706cbf2ce58e63ab8bbc9d16453b2c792ed4 is empty > error: unable to find 93d7706cbf2ce58e63ab8bbc9d16453b2c792ed4 > fatal: SHA1 COLLISION FOUND WITH 93d7706cbf2ce58e63ab8bbc9d16453b2c792ed4 ! > fatal: index-pack failed > Your configuration specifies to merge with the ref 'master' > from the remote, but no such ref was fetched. > > I don't see any problems on my machine Since I posted the above some animals have built successfully, so it seems to be only affecting some animals. -- Kevin Grittner EDB: http://www.enterprisedb.com The Enterprise PostgreSQL Company
On 2015-01-30 15:34:02 +0000, Kevin Grittner wrote: > My push to branches 9.3 and up seems to have broken the buildfarm > with this: I think this isn't anything related to your commit. Both racoon and macaque have failed for a while. They just happened to run quickly after your commit, giving the impression that two animals failed consecutively. Greetings, Andres Freund -- Andres Freund http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training & Services
Kevin,
* Kevin Grittner (kgrittn@ymail.com) wrote:
> My push to branches 9.3 and up seems to have broken the buildfarm
> with this:
I don't think it has anything to do with your push. A number of members
have built with your latest and look fine at this point, and I'm not
seeing any issues here either.
I do see that macque is having issues, but it's been busted for the past
3 days it looks like.
We have backups from prior to your commit, just in case, but I don't
think there's anything wrong.
Thanks!
Stephen
Stephen Frost <sfrost@snowman.net> writes:
> * Kevin Grittner (kgrittn@ymail.com) wrote:
>> My push to branches 9.3 and up seems to have broken the buildfarm
>> with this:
> I don't think it has anything to do with your push. A number of members
> have built with your latest and look fine at this point, and I'm not
> seeing any issues here either.
> I do see that macque is having issues, but it's been busted for the past
> 3 days it looks like.
Yeah. Several of the critters have been having git issues for weeks
to months. I wonder whether we couldn't teach the buildfarm script
to recover from this automatically ...
regards, tom lane
On 01/30/2015 11:00 AM, Tom Lane wrote: > >> I do see that macque is having issues, but it's been busted for the past >> 3 days it looks like. > Yeah. Several of the critters have been having git issues for weeks > to months. I wonder whether we couldn't teach the buildfarm script > to recover from this automatically ... > > Could get very messy, especially when the error is in the mirror, as is apparently the case here. As I mentioned the other day on the buildfarm list, I'm playing with a mode that would largely obviate the need for a mirror. Failure cases are what's worrying me, though. In this particular case, the owner should probably remove the mirror and each of the branch pgsql directories. cheers andrew