Re: Improve TAP tests of pg_upgrade for cross-version tests
От | Andrew Dunstan |
---|---|
Тема | Re: Improve TAP tests of pg_upgrade for cross-version tests |
Дата | |
Msg-id | c3198cb1-fe79-a92b-565a-cf1fd3c1c01d@dunslane.net обсуждение исходный текст |
Ответ на | Re: Improve TAP tests of pg_upgrade for cross-version tests (Michael Paquier <michael@paquier.xyz>) |
Список | pgsql-hackers |
On 2022-06-13 Mo 03:51, Michael Paquier wrote: > On Sun, Jun 12, 2022 at 05:58:54PM -0400, Andrew Dunstan wrote: >> On 2022-06-12 Su 10:14, Andrew Dunstan wrote: >>> I tried in fb16d2c658 to avoid littering the mainline code with >>> version-specific tests, and put that in the methods in the subclasses >>> that override the mainline functions. > Except that manipulating the diffs of pg_upgrade is not something that > needs to be internal to the subclasses where we set up the nodes :) > >> I think I must have been insufficiently caffeinated when I wrote this, >> because clearly I was not reading correctly. >> >> I have had another look and the patch looks fine, although I haven't >> tested it. > I have a question about the tests done in the buildfarm though. Do > the dumps from the older versions drop some of the objects that cause > the diffs in the dumps? At which extent is that a dump from > installcheck? See lines 324..347 (save stage) and 426..586 (upgrade stage) of <https://github.com/PGBuildFarm/client-code/blob/main/PGBuild/Modules/TestUpgradeXversion.pm> We save the cluster to be upgraded after all the installcheck stages have run, so on crake here's the list of databases upgraded for HEAD: postgres template1 template0 contrib_regression_pgxml contrib_regression_bool_plperl contrib_regression_hstore_plperl contrib_regression_jsonb_plperl regression contrib_regression_redis_fdw contrib_regression_file_textarray_fdw isolation_regression pl_regression_plpgsql contrib_regression_hstore_plpython3 pl_regression_plperl pl_regression_plpython3 pl_regression_pltcl contrib_regression_adminpack contrib_regression_amcheck contrib_regression_bloom contrib_regression_btree_gin contrib_regression_btree_gist contrib_regression_citext contrib_regression_cube contrib_regression_dblink contrib_regression_dict_int contrib_regression_dict_xsyn contrib_regression_earthdistance contrib_regression_file_fdw contrib_regression_fuzzystrmatch contrib_regression_hstore contrib_regression__int contrib_regression_isn contrib_regression_lo contrib_regression_ltree contrib_regression_pageinspect contrib_regression_passwordcheck contrib_regression_pg_surgery contrib_regression_pg_trgm contrib_regression_pgstattuple contrib_regression_pg_visibility contrib_regression_postgres_fdw contrib_regression_seg contrib_regression_tablefunc contrib_regression_tsm_system_rows contrib_regression_tsm_system_time contrib_regression_unaccent contrib_regression_pgcrypto contrib_regression_uuid-ossp contrib_regression_jsonb_plpython3 contrib_regression_ltree_plpython3 isolation_regression_summarization-and-inprogress-insertion isolation_regression_delay_execution contrib_regression_dummy_index_am contrib_regression_dummy_seclabel contrib_regression_plsample contrib_regression_spgist_name_ops contrib_regression_test_bloomfilter contrib_regression_test_extensions contrib_regression_test_ginpostinglist contrib_regression_test_integerset contrib_regression_test_parser contrib_regression_test_pg_dump contrib_regression_test_predtest contrib_regression_test_rbtree contrib_regression_test_regex contrib_regression_test_rls_hooks contrib_regression_test_shm_mq contrib_regression_rolenames cheers andrew -- Andrew Dunstan EDB: https://www.enterprisedb.com
В списке pgsql-hackers по дате отправления:
Предыдущее
От: Nitin JadhavДата:
Сообщение: Re: Report checkpoint progress with pg_stat_progress_checkpoint (was: Report checkpoint progress in server logs)
Следующее
От: Tomas VondraДата:
Сообщение: Re: pgcon unconference / impact of block size on performance