Re: Teach pg_upgrade test to honor NO_TEMP_INSTALL

Поиск
Список
Период
Сортировка
От Andrew Dunstan
Тема Re: Teach pg_upgrade test to honor NO_TEMP_INSTALL
Дата
Msg-id 2b42bb13-7a7e-e295-2054-9a209749c3a7@2ndQuadrant.com
обсуждение исходный текст
Ответ на Re: Teach pg_upgrade test to honor NO_TEMP_INSTALL  (Andres Freund <andres@anarazel.de>)
Ответы Re: Teach pg_upgrade test to honor NO_TEMP_INSTALL  (Andres Freund <andres@anarazel.de>)
Список pgsql-hackers
On 5/22/19 2:42 PM, Andres Freund wrote:
> Hi,
>
> On 2019-05-22 14:27:51 -0400, Tom Lane wrote:
>> Andres Freund <andres@anarazel.de> writes:
>>> On 2019-05-22 14:06:47 -0400, Tom Lane wrote:
>>>> Not sure about that last bit.  pg_upgrade has the issue of possibly
>>>> wanting to deal with 2 installations, unlike the rest of the tree,
>>>> so I'm not sure that fixing its problem means there's something we
>>>> need to change everywhere else.
>>> I'm not quite following? We need to move it into global scope to fix the
>>> issue at hand (namely that we currently need to make install first, just
>>> to get psql).  And at which scope could it be in master, other than
>>> global?
>> Maybe I misunderstood you --- I thought you were talking about something
>> like defining EXTRA_REGRESS_OPTS in Makefile.global.  If you mean
>> running this unconditionally within test.sh, I've got no objection
>> to that.
> Oh, yes, that's what I meant.
>
>
>>> I do think we will have to move it to the global scope in the back
>>> branches too, because NO_TEMP_INSTALL does indeed fail without a global
>>> install first (rather than using the temp install, as intended):
>> Agreed, we should fix it in all branches, because it seems like it's
>> probably testing the wrong thing, ie using the later branch's psql
>> to run the earlier branch's regression tests.



If I disable install, the buildfarm fails the upgrade check even when
not using NO_TEMP_INSTALL.


excerpts from the log:



rm -rf '/home/pgl/npgl/pg_head/bfroot/HEAD/pgsql.build'/tmp_install
/bin/mkdir -p
'/home/pgl/npgl/pg_head/bfroot/HEAD/pgsql.build'/tmp_install/log
make -C '../../..'
DESTDIR='/home/pgl/npgl/pg_head/bfroot/HEAD/pgsql.build'/tmp_install
install
>'/home/pgl/npgl/pg_head/bfroot/HEAD/pgsql.build'/tmp_install/log/install.log
2>&1
make -j1  checkprep
>>'/home/pgl/npgl/pg_head/bfroot/HEAD/pgsql.build'/tmp_install/log/install.log
2>&1
MAKE=make
PATH="/home/pgl/npgl/pg_head/bfroot/HEAD/pgsql.build/tmp_install/home/pgl/npgl/pg_head/bfroot/HEAD/inst/bin:$PATH"

LD_LIBRARY_PATH="/home/pgl/npgl/pg_head/bfroot/HEAD/pgsql.build/tmp_install/home/pgl/npgl/pg_head/bfroot/HEAD/inst/lib" 
bindir=/home/pgl/npgl/pg_h
ead/bfroot/HEAD/pgsql.build/tmp_install//home/pgl/npgl/pg_head/bfroot/HEAD/inst/bin
EXTRA_REGRESS_OPTS="--port=5678" /bin/sh
/home/pgl/npgl/pg_head/src/bin/pg_upgrade/test.sh


rm -rf ./testtablespace
mkdir ./testtablespace
../../../src/test/regress/pg_regress
--inputdir=/home/pgl/npgl/pg_head/src/test/regress
--bindir='/home/pgl/npgl/pg_head/bfroot/HEAD/inst/bin'   --port=5678
--port=54464 --dlpath=. --max-concurrent-tests=20 --port=5678
--port=54464
--schedule=/home/pgl/npgl/pg_head/src/test/regress/parallel_schedule 
(using postmaster on /tmp/pg_upgrade_check-GCUkGu, port 54464)
============== dropping database "regression"         ==============
sh: /home/pgl/npgl/pg_head/bfroot/HEAD/inst/bin/psql: No such file or
directory
command failed: "/home/pgl/npgl/pg_head/bfroot/HEAD/inst/bin/psql" -X -c
"DROP DATABASE IF EXISTS \"regression\"" "postgres"
make[1]: *** [GNUmakefile:141: installcheck-parallel] Error 2
make[1]: Leaving directory
'/home/pgl/npgl/pg_head/bfroot/HEAD/pgsql.build/src/test/regress'
make: *** [GNUmakefile:68: installcheck-parallel] Error 2
make: Leaving directory '/home/pgl/npgl/pg_head/bfroot/HEAD/pgsql.build'
waiting for server to shut down.... done
server stopped
make: *** [Makefile:48: check] Error 1



It looks to me like the bindir needs to be passed to the make called by
test.sh (maybe LD_LIBRARY_PATH too?)


cheers


andrew


-- 

Andrew Dunstan                https://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services




В списке pgsql-hackers по дате отправления:

Предыдущее
От: Donald Dong
Дата:
Сообщение: Re: Why could GEQO produce plans with lower costs than thestandard_join_search?
Следующее
От: Andres Freund
Дата:
Сообщение: Re: Teach pg_upgrade test to honor NO_TEMP_INSTALL