Adding +1 to fix upgrade path from F26 -> F27.
Based on the chat during our meeting; if you want to guarantee that copr-backend goes into Fedora at the same time when copr-rpmbuild goes, you should have both packages shipped through the same bodhi update.
mock chroot config for fedora 27 is not yet in F26, blocked by this ... could the configuration have separate life-cycle (in separate package)?
Fixes the sclo-postgresql96-postgresql build for me. Thanks!
mea culpa :( I've missed that the python packages are in noarch category, I'm sorry ... but I'm still seeing:
$ file /tmp/postgres-9.2.mo
/tmp/postgres-9.2.mo: new-fs dump file (little endian), Previous dump Thu Jan 1 00:00:00 1970, This dump Thu Jan 1 01:02:29 1970, Volume 28, Level 33, type: Label ", Filesystem PG\001, Device MH\001, Host \027J\001, Flags 14b5a
$ rpm -q file file-libs python3-magic python-magic
file-5.30-7.fc26.x86_64
file-libs-5.30-7.fc26.x86_64
python3-magic-5.30-7.fc26.noarch
python-magic-5.30-7.fc26.noarch
Seems like this is not bug in package itself -- but the update is broken: Tthe packages are built (based on root.log) but koji (for some reason) did not add python-magic and python3-magic packages into the resultdir:
$ lftp https://kojipkgs.fedoraproject.org//packages/file/5.30/7.fc26/x86_64/
cd ok, cwd=//packages/file/5.30/7.fc26/x86_64
lftp kojipkgs.fedoraproject.org://packages/file/5.30/7.fc26/x86_64> ls
drwxr-xr-x -- /packages/file/5.30/7.fc26
-rw-r--r-- 70K 2017-07-26 07:54 file-5.30-7.fc26.x86_64.rpm
-rw-r--r-- 256K 2017-07-26 07:54 file-debuginfo-5.30-7.fc26.x86_64.rpm
-rw-r--r-- 39K 2017-07-26 07:54 file-devel-5.30-7.fc26.x86_64.rpm
-rw-r--r-- 504K 2017-07-26 07:54 file-libs-5.30-7.fc26.x86_64.rpm
lftp kojipkgs.fedoraproject.org://packages/file/5.30/7.fc26/x86_64>
Thanks for the quick fix!
I was able to build PostgreSQL against js 1:1.8.5-27.fc27 in rawhide, thanks!
Works for me now. Thanks!
The 'sources' file generated by 'fedpkg upload' is not readable by koji? https://kojipkgs.fedoraproject.org//work/tasks/707/16850707/mock_output.log
Just a test-case fix.
False alarm, see bug 1349381 for more info.
Seems like this breaks automake's testsuite (FTBFS), I don't want to give it negative karma -- but if you plan to push the update to stable, please allow me some time to debug properly.
Works for me.
[root@test-postgresql-22 dtf-postgresql]# ./DTF_OUTPUT/initdb/run
initdb/basic -> OK
** Success. **
[root@test-postgresql-22 dtf-postgresql]# ./DTF_OUTPUT/upgrade/run
upgrade/basic -> OK
** going to upgrade/locale **
upgrade/locale/changed -> OK
upgrade/locale/utf8-syntax -> OK
** group upgrade/locale: Ok. **
** Success. **
[root@test-postgresql-22 dtf-postgresql]# ./DTF_OUTPUT/upstream-testsuite/run
upstream-testsuite -> OK
** Success. **
[root@test-postgresql-22 dtf-postgresql]# rpm -q postgresql-server postgresql-test
postgresql-server-9.4.6-1.fc22.x86_64
postgresql-test-9.4.6-1.fc22.x86_64
[root@test-postgresql-23 dtf-postgresql]# ./DTF_OUTPUT/upgrade/run
upgrade/basic -> OK
** going to upgrade/locale **
upgrade/locale/changed -> OK
upgrade/locale/utf8-syntax -> OK
** group upgrade/locale: Ok. **
** Success. **
[root@test-postgresql-23 dtf-postgresql]# ./DTF_OUTPUT/initdb/run
initdb/basic -> OK
** Success. **
[root@test-postgresql-23 dtf-postgresql]# rpm -q postgresql-server
postgresql-server-9.4.6-1.fc23.x86_64
works for mi in fedora-22-x86_64 chroot
Thanks, tested on: https://github.com/dhodovsk/postgresql-container/tree/distgen