[Tarantool-patches] [PATCH v3] replication: the truncate method called from within a transaction

Serge Petrenko sergepetrenko at tarantool.org
Wed Nov 3 17:37:29 MSK 2021



03.11.2021 13:03, Ян Штундер пишет:
> Thanks for the comments!
>  I corrected your remarks.

Thanks for the changes!
LGTM now.

>
> Diff:
>
> diff --git 
> a/test/replication-luatest/gh_6123_truncate_temp_space_test.lua 
> b/test/replication-luatest/gh_6123_truncate_temp_space_test.lua
> new file mode 100644
> index 000000000..7940e722f
> --- /dev/null
> +++ b/test/replication-luatest/gh_6123_truncate_temp_space_test.lua
> @@ -0,0 +1,62 @@
> +local t = require('luatest')
> +local cluster = require('test.luatest_helpers.cluster')
> +local helpers = require('test.luatest_helpers')
> +
> +local g = t.group('gh_6123', {{engine = 'memtx'}, {engine = 'vinyl'}})
> +
> +g.before_each(function(cg)
> +    local engine = cg.params.engine
> +
> +    cg.cluster = cluster:new({})
> +
> +    local box_cfg = {
> +        replication         = {
> +            helpers.instance_uri('master')
> +        },
> +        replication_timeout = 1,
> +        read_only           = false
> +    }
> +
> +    cg.master = cg.cluster:build_server({alias = 'master', engine = 
> engine, box_cfg = box_cfg})
> +
> +    local box_cfg = {
> +        replication         = {
> +            helpers.instance_uri('master'),
> +            helpers.instance_uri('replica')
> +        },
> +        replication_timeout = 1,
> +        replication_connect_timeout = 4,
> +        read_only           = true
> +    }
> +
> +    cg.replica = cg.cluster:build_server({alias = 'replica', engine = 
> engine, box_cfg = box_cfg})
> +
> +    cg.cluster:add_server(cg.master)
> +    cg.cluster:add_server(cg.replica)
> +    cg.cluster:start()
> +end)
> +
> +
> +g.after_each(function(cg)
> +    cg.cluster.servers = nil
> +    cg.cluster:drop()
> +end)
> +
> +
> +g.test_truncate_is_local_transaction = function(cg)
> +    cg.master:eval("s = box.schema.space.create('temp', {temporary = 
> true})")
> +    cg.master:eval("s:create_index('pk')")
> +
> +    cg.master:eval("s:insert{1, 2}")
> +    cg.master:eval("s:insert{4}")
> +    t.assert_equals(cg.master:eval("return s:select()"), {{1, 2}, {4}})
> +
> +    cg.master:eval("box.begin() box.space._schema:replace{'smth'} 
> s:truncate() box.commit()")
> +    t.assert_equals(cg.master:eval("return s:select()"), {})
> +    t.assert_equals(cg.master:eval("return 
> box.space._schema:select{'smth'}"), {{'smth'}})
> +
> +    -- Checking that replica has received the last transaction,
> +    -- and that replication isn't broken.
> +    t.assert_equals(cg.replica:eval("return 
> box.space._schema:select{'smth'}"), {{'smth'}})
> +    t.assert_equals(cg.replica:eval("return 
> box.info.replication[1].upstream.status"), 'follow')
> +end
>
> --
> Yan Shtunder
>
> пт, 29 окт. 2021 г. в 12:20, Serge Petrenko <sergepetrenko at tarantool.org>:
>
>
>
>     28.10.2021 18:48, Ян Штундер пишет:
>     > Hi! Thank you for the review!
>     > I have fixed the errors
>     >
>     >     Please, find a better test name.
>     >     Like "gh_6123_truncate_temp_space_test.lua"
>     >
>     >
>     > gh_6123_test.lua -> gh_6123_truncate_temp_space_test.lua
>     >
>     > --
>     > Yan Shtunder
>     >
>     >
>
>     Thanks  for the fixes!
>
>     One comment left. I'm pasting part of the patch here to review.
>
>     In test:
>
>     +g.after_each(function(cg)
>     +    cg.cluster.servers = nil
>     +    cg.cluster:drop()
>     +    fio.rmtree(cg.master.workdir)
>     +    fio.rmtree(cg.replica.workdir)
>     +end)
>     +
>     +
>     +g.before_test('test_truncate_is_local_transaction', function(cg)
>     +    cg.cluster:add_server(cg.master)
>     +    cg.cluster:add_server(cg.replica)
>     +    cg.cluster:start()
>     +end)
>     +
>     +g.after_test('test_truncate_is_local_transaction', function(cg)
>     +    cg.cluster:drop({cg.master, cg.replica})
>     +end)
>     +
>
>     You have a single test: test_truncate_is_local_transaction.
>
>     So g.after_each and g.after_test both have the same meaning in
>     your case.
>
>     I propose to dorp g.after_test, since g.after_each supersedes it.
>
>     Also, you don't need to call fio.rmtree in g.after_each.
>     luatest does this for you, when needed.
>     >
>     > пн, 25 окт. 2021 г. в 16:34, Serge Petrenko
>     <sergepetrenko at tarantool.org>:
>     >
>     >
>     >
>     >     25.10.2021 13:44, Yan Shtunder via Tarantool-patches пишет:
>     >     > The truncate method could be called from within a transaction.
>     >     The flag
>     >     > of GROUP_LOCAL was set in truncate method after statement row
>     >     had been
>     >     > being checked on the GROUP_LOCAL. Accordingly, after a local
>     >     transaction
>     >     > NOP row was not appended.
>     >     >
>     >     > Closes #6123
>     >
>     >     Thanks for the patch!
>     >
>     >     Please, find 2 comments below.
>     >     > ---
>     >     > Issue: https://github.com/tarantool/tarantool/issues/6123
>     >     > Patch:
>     >
>     https://github.com/tarantool/tarantool/tree/yshtunder/gh-6123-truncate-is-local-transaction
>     >     >
>     >     >   src/box/alter.cc                          | 5 ++
>     >     >   test/replication-luatest/gh_6123_test.lua | 62
>     >     +++++++++++++++++++++++
>     >     >   2 files changed, 67 insertions(+)
>     >     >   create mode 100644 test/replication-luatest/gh_6123_test.lua
>     >     >
>     >     > diff --git a/src/box/alter.cc b/src/box/alter.cc
>     >     > index 31ac82fc5..52b9ea4d2 100644
>     >     > --- a/src/box/alter.cc
>     >     > +++ b/src/box/alter.cc
>     >     > @@ -2914,6 +2914,11 @@ on_replace_dd_truncate(struct trigger *
>     >     /* trigger */, void *event)
>     >     >       if (space_is_temporary(old_space) ||
>     >     >           space_group_id(old_space) == GROUP_LOCAL) {
>     >     >               stmt->row->group_id = GROUP_LOCAL;
>     >     > +             /*
>     >     > +              * The trigger is invoked after
>     txn->n_local_rows
>     >     > +              * is counted, so don't forget to update it
>     here.
>     >     > +              */
>     >     > +             ++txn->n_local_rows;
>     >     >       }
>     >     >
>     >     >       try {
>     >
>     >     The patch doesn't build on my machine. For the same reason
>     as the
>     >     other
>     >     patch.
>     >     I suppose you'll fix that with the other patch.
>     >
>     >     > diff --git a/test/replication-luatest/gh_6123_test.lua
>     >     b/test/replication-luatest/gh_6123_test.lua
>     >     > new file mode 100644
>     >     > index 000000000..d0ffa0c8b
>     >     > --- /dev/null
>     >     > +++ b/test/replication-luatest/gh_6123_test.lua
>     >
>     >     Please, find a better test name.
>     >     Like "gh_6123_truncate_temp_space_test.lua"
>     >     > @@ -0,0 +1,62 @@
>     >     > +local fio = require('fio')
>     >     > +local log = require('log')
>     >     > +local t = require('luatest')
>     >     > +local cluster = require('test.luatest_helpers.cluster')
>     >     > +local helpers = require('test.luatest_helpers.helpers')
>     >     > +
>     >     > +local g = t.group('gh-6123')
>     >     > +
>     >     > +g.before_test('test_truncate_is_local_transaction',
>     function()
>     >     > +    g.cluster = cluster:new({})
>     >     > +
>     >     > +    local box_cfg = {
>     >     > +        replication         = {
>     >     > +            helpers.instance_uri('master')
>     >     > +        },
>     >     > +        replication_timeout = 0.1,
>     >     > +        read_only           = false
>     >     > +    }
>     >     > +
>     >     > +    g.master = g.cluster:build_server({alias = 'master'},
>     >     engine, box_cfg)
>     >     > +
>     >     > +    local box_cfg = {
>     >     > +        replication         = {
>     >     > +            helpers.instance_uri('master'),
>     >     > +            helpers.instance_uri('replica')
>     >     > +        },
>     >     > +        replication_timeout = 0.1,
>     >     > +        replication_connect_timeout = 0.5,
>     >     > +        read_only           = true
>     >     > +    }
>     >     > +
>     >     > +    g.replica = g.cluster:build_server({alias = 'replica'},
>     >     engine, box_cfg)
>     >     > +
>     >     > +    g.cluster:join_server(g.master)
>     >     > +    g.cluster:join_server(g.replica)
>     >     > +    g.cluster:start()
>     >     > + log.info <http://log.info> <http://log.info>('Everything
>     is started')
>     >     > +end)
>     >     > +
>     >     > +g.after_test('test_truncate_is_local_transaction', function()
>     >     > +    g.cluster:stop()
>     >     > +    fio.rmtree(g.master.workdir)
>     >     > +    fio.rmtree(g.replica.workdir)
>     >     > +end)
>     >     > +
>     >     > +g.test_truncate_is_local_transaction = function()
>     >     > +    g.master:eval("s = box.schema.space.create('temp',
>     >     {temporary = true})")
>     >     > +    g.master:eval("s:create_index('pk')")
>     >     > +
>     >     > +    g.master:eval("s:insert{1, 2}")
>     >     > +    g.master:eval("s:insert{4}")
>     >     > +    t.assert_equals(g.master:eval("return s:select()"), {{1,
>     >     2}, {4}})
>     >     > +
>     >     > +    g.master:eval("box.begin()
>     >     box.space._schema:replace{'smth'} s:truncate() box.commit()")
>     >     > +    t.assert_equals(g.master:eval("return s:select()"), {})
>     >     > +    t.assert_equals(g.master:eval("return
>     >     box.space._schema:select{'smth'}"), {{'smth'}})
>     >     > +
>     >     > +    -- Checking that replica has received the last
>     transaction,
>     >     > +    -- and that replication isn't broken.
>     >     > +    t.assert_equals(g.replica:eval("return
>     >     box.space._schema:select{'smth'}"), {{'smth'}})
>     >     > +    t.assert_equals(g.replica:eval("return
>     >     box.info.replication[1].upstream.status"), 'follow')
>     >     > +end
>     >     > --
>     >     > 2.25.1
>     >     >
>     >
>     >     --
>     >     Serge Petrenko
>     >
>
>     -- 
>     Serge Petrenko
>

-- 
Serge Petrenko



More information about the Tarantool-patches mailing list