Re: [Tarantool-patches] [PATCH v2 20/19] replication: add test for quorum 1

[Serge Petrenko <sergepetrenko@tarantool.org>]

01.07.2020 02:00, Vladislav Shpilevoy пишет:
> When synchro quorum is 1, the final commit and confirmation write
> are done by the fiber created the transaction, right after WAL
> write. This case got special handling in the previous patches,
> and this commits adds a test for that.
>
> Closes #5123


Thanks for the patch! LGTM.

> ---
>   test/replication/qsync_basic.result    |  33 +++++++
>   test/replication/qsync_basic.test.lua  |  12 +++
>   test/replication/qsync_errinj.result   | 114 +++++++++++++++++++++++++
>   test/replication/qsync_errinj.test.lua |  45 ++++++++++
>   test/replication/suite.ini             |   2 +-
>   5 files changed, 205 insertions(+), 1 deletion(-)
>   create mode 100644 test/replication/qsync_errinj.result
>   create mode 100644 test/replication/qsync_errinj.test.lua
>
> diff --git a/test/replication/qsync_basic.result b/test/replication/qsync_basic.result
> index f713d4b08..cdecf00e8 100644
> --- a/test/replication/qsync_basic.result
> +++ b/test/replication/qsync_basic.result
> @@ -299,6 +299,39 @@ box.space.sync:select{6}
>    | - []
>    | ...
>   
> +--
> +-- gh-5123: quorum 1 still should write CONFIRM.
> +--
> +test_run:switch('default')
> + | ---
> + | - true
> + | ...
> +box.cfg{replication_synchro_quorum = 1, replication_synchro_timeout = 5}
> + | ---
> + | ...
> +oldlsn = box.info.lsn
> + | ---
> + | ...
> +box.space.sync:replace{7}
> + | ---
> + | - [7]
> + | ...
> +newlsn = box.info.lsn
> + | ---
> + | ...
> +assert(newlsn >= oldlsn + 2)
> + | ---
> + | - true
> + | ...
> +test_run:switch('replica')
> + | ---
> + | - true
> + | ...
> +box.space.sync:select{7}
> + | ---
> + | - - [7]
> + | ...
> +
>   -- Cleanup.
>   test_run:cmd('switch default')
>    | ---
> diff --git a/test/replication/qsync_basic.test.lua b/test/replication/qsync_basic.test.lua
> index f84b6ee19..361f22bc3 100644
> --- a/test/replication/qsync_basic.test.lua
> +++ b/test/replication/qsync_basic.test.lua
> @@ -118,6 +118,18 @@ test_run:switch('replica')
>   box.space.test:select{6}
>   box.space.sync:select{6}
>   
> +--
> +-- gh-5123: quorum 1 still should write CONFIRM.
> +--
> +test_run:switch('default')
> +box.cfg{replication_synchro_quorum = 1, replication_synchro_timeout = 5}
> +oldlsn = box.info.lsn
> +box.space.sync:replace{7}
> +newlsn = box.info.lsn
> +assert(newlsn >= oldlsn + 2)
> +test_run:switch('replica')
> +box.space.sync:select{7}
> +
>   -- Cleanup.
>   test_run:cmd('switch default')
>   
> diff --git a/test/replication/qsync_errinj.result b/test/replication/qsync_errinj.result
> new file mode 100644
> index 000000000..1d2945761
> --- /dev/null
> +++ b/test/replication/qsync_errinj.result
> @@ -0,0 +1,114 @@
> +-- test-run result file version 2
> +test_run = require('test_run').new()
> + | ---
> + | ...
> +engine = test_run:get_cfg('engine')
> + | ---
> + | ...
> +
> +old_synchro_quorum = box.cfg.replication_synchro_quorum
> + | ---
> + | ...
> +old_synchro_timeout = box.cfg.replication_synchro_timeout
> + | ---
> + | ...
> +box.schema.user.grant('guest', 'super')
> + | ---
> + | ...
> +
> +test_run:cmd('create server replica with rpl_master=default,\
> +             script="replication/replica.lua"')
> + | ---
> + | - true
> + | ...
> +test_run:cmd('start server replica with wait=True, wait_load=True')
> + | ---
> + | - true
> + | ...
> +
> +_ = box.schema.space.create('sync', {is_sync = true, engine = engine})
> + | ---
> + | ...
> +_ = box.space.sync:create_index('pk')
> + | ---
> + | ...
> +
> +--
> +-- gh-5123: replica WAL fail shouldn't crash with quorum 1.
> +--
> +test_run:switch('default')
> + | ---
> + | - true
> + | ...
> +box.cfg{replication_synchro_quorum = 1, replication_synchro_timeout = 5}
> + | ---
> + | ...
> +box.space.sync:insert{1}
> + | ---
> + | - [1]
> + | ...
> +
> +test_run:switch('replica')
> + | ---
> + | - true
> + | ...
> +box.error.injection.set('ERRINJ_WAL_IO', true)
> + | ---
> + | - ok
> + | ...
> +
> +test_run:switch('default')
> + | ---
> + | - true
> + | ...
> +box.space.sync:insert{2}
> + | ---
> + | - [2]
> + | ...
> +
> +test_run:switch('replica')
> + | ---
> + | - true
> + | ...
> +test_run:wait_upstream(1, {status='stopped'})
> + | ---
> + | - true
> + | ...
> +box.error.injection.set('ERRINJ_WAL_IO', false)
> + | ---
> + | - ok
> + | ...
> +
> +test_run:cmd('restart server replica')
> + |
> +box.space.sync:select{2}
> + | ---
> + | - - [2]
> + | ...
> +
> +test_run:cmd('switch default')
> + | ---
> + | - true
> + | ...
> +
> +box.cfg{                                                                        \
> +    replication_synchro_quorum = old_synchro_quorum,                            \
> +    replication_synchro_timeout = old_synchro_timeout,                          \
> +}
> + | ---
> + | ...
> +test_run:cmd('stop server replica')
> + | ---
> + | - true
> + | ...
> +test_run:cmd('delete server replica')
> + | ---
> + | - true
> + | ...
> +
> +box.space.sync:drop()
> + | ---
> + | ...
> +box.schema.user.revoke('guest', 'super')
> + | ---
> + | ...
> diff --git a/test/replication/qsync_errinj.test.lua b/test/replication/qsync_errinj.test.lua
> new file mode 100644
> index 000000000..96495ae6c
> --- /dev/null
> +++ b/test/replication/qsync_errinj.test.lua
> @@ -0,0 +1,45 @@
> +test_run = require('test_run').new()
> +engine = test_run:get_cfg('engine')
> +
> +old_synchro_quorum = box.cfg.replication_synchro_quorum
> +old_synchro_timeout = box.cfg.replication_synchro_timeout
> +box.schema.user.grant('guest', 'super')
> +
> +test_run:cmd('create server replica with rpl_master=default,\
> +             script="replication/replica.lua"')
> +test_run:cmd('start server replica with wait=True, wait_load=True')
> +
> +_ = box.schema.space.create('sync', {is_sync = true, engine = engine})
> +_ = box.space.sync:create_index('pk')
> +
> +--
> +-- gh-5123: replica WAL fail shouldn't crash with quorum 1.
> +--
> +test_run:switch('default')
> +box.cfg{replication_synchro_quorum = 1, replication_synchro_timeout = 5}
> +box.space.sync:insert{1}
> +
> +test_run:switch('replica')
> +box.error.injection.set('ERRINJ_WAL_IO', true)
> +
> +test_run:switch('default')
> +box.space.sync:insert{2}
> +
> +test_run:switch('replica')
> +test_run:wait_upstream(1, {status='stopped'})
> +box.error.injection.set('ERRINJ_WAL_IO', false)
> +
> +test_run:cmd('restart server replica')
> +box.space.sync:select{2}
> +
> +test_run:cmd('switch default')
> +
> +box.cfg{                                                                        \
> +    replication_synchro_quorum = old_synchro_quorum,                            \
> +    replication_synchro_timeout = old_synchro_timeout,                          \
> +}
> +test_run:cmd('stop server replica')
> +test_run:cmd('delete server replica')
> +
> +box.space.sync:drop()
> +box.schema.user.revoke('guest', 'super')
> diff --git a/test/replication/suite.ini b/test/replication/suite.ini
> index 6119a264b..11f8d4e20 100644
> --- a/test/replication/suite.ini
> +++ b/test/replication/suite.ini
> @@ -3,7 +3,7 @@ core = tarantool
>   script =  master.lua
>   description = tarantool/box, replication
>   disabled = consistent.test.lua
> -release_disabled = catch.test.lua errinj.test.lua gc.test.lua gc_no_space.test.lua before_replace.test.lua quorum.test.lua recover_missing_xlog.test.lua sync.test.lua long_row_timeout.test.lua gh-4739-vclock-assert.test.lua gh-4730-applier-rollback.test.lua
> +release_disabled = catch.test.lua errinj.test.lua gc.test.lua gc_no_space.test.lua before_replace.test.lua qsync_errinj.test.lua quorum.test.lua recover_missing_xlog.test.lua sync.test.lua long_row_timeout.test.lua gh-4739-vclock-assert.test.lua gh-4730-applier-rollback.test.lua
>   config = suite.cfg
>   lua_libs = lua/fast_replica.lua lua/rlimit.lua
>   use_unix_sockets = True

-- 
Serge Petrenko