[Tarantool-patches] [PATCH 2/2] replication: support ROLLBACK and CONFIRM during recovery

[Serge Petrenko]

19.06.2020 21:00, Serge Petrenko пишет:
> Follow-up #4847
> Follow-up #4848

Added a small test for CONFIRM/ROLLBACK:

commit 5b91fbd3996c4ca43f5c96a1767e5b7ec7cea86d
Author: Serge Petrenko <sergepetrenko at tarantool.org>
Date:   Tue Jun 23 14:20:45 2020 +0300

     replication: add test for synchro CONFIRM/ROLLBACK

     [TO BE SQUASHED INTO THE PREVIOUS COMMIT]

diff --git a/test/replication/sync_replication_sanity.result 
b/test/replication/sync_replication_sanity.result
index 551df7daf..4b9823d77 100644
--- a/test/replication/sync_replication_sanity.result
+++ b/test/replication/sync_replication_sanity.result
@@ -69,3 +69,135 @@ box.schema.create_space('test', {is_sync = true, 
is_local = true})
   | ---
   | - error: 'Failed to create space ''test'': local space can''t be 
synchronous'
   | ...
+
+--
+-- gh-4847, gh-4848: CONFIRM and ROLLBACK entries in WAL.
+--
+env = require('test_run')
+ | ---
+ | ...
+test_run = env.new()
+ | ---
+ | ...
+fiber = require('fiber')
+ | ---
+ | ...
+engine = test_run:get_cfg('engine')
+ | ---
+ | ...
+
+box.schema.user.grant('guest', 'replication')
+ | ---
+ | ...
+-- Set up synchronous replication options.
+quorum = box.cfg.replication_synchro_quorum
+ | ---
+ | ...
+timeout = box.cfg.replication_synchro_timeout
+ | ---
+ | ...
+box.cfg{replication_synchro_quorum=2, replication_synchro_timeout=0.1}
+ | ---
+ | ...
+
+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')
+ | ---
+ | ...
+
+lsn = box.info.lsn
+ | ---
+ | ...
+box.space.sync:insert{1}
+ | ---
+ | - [1]
+ | ...
+-- 1 for insertion, 1 for CONFIRM message.
+box.info.lsn - lsn
+ | ---
+ | - 2
+ | ...
+-- Raise quorum so that master has to issue a ROLLBACK.
+box.cfg{replication_synchro_quorum=3}
+ | ---
+ | ...
+t = fiber.time()
+ | ---
+ | ...
+box.space.sync:insert{2}
+ | ---
+ | - error: Quorum collection for a synchronous transaction is timed out
+ | ...
+-- Check that master waited for acks.
+fiber.time() - t > box.cfg.replication_synchro_timeout
+ | ---
+ | - true
+ | ...
+box.cfg{replication_synchro_quorum=2}
+ | ---
+ | ...
+box.space.sync:insert{3}
+ | ---
+ | - [3]
+ | ...
+box.space.sync:select{}
+ | ---
+ | - - [1]
+ |   - [3]
+ | ...
+
+-- Check consistency on replica.
+test_run:cmd('switch replica')
+ | ---
+ | - true
+ | ...
+box.space.sync:select{}
+ | ---
+ | - - [1]
+ |   - [3]
+ | ...
+
+-- Check consistency in recovered data.
+test_run:cmd('restart server replica')
+ |
+box.space.sync:select{}
+ | ---
+ | - - [1]
+ |   - [3]
+ | ...
+
+-- Cleanup.
+test_run:cmd('switch default')
+ | ---
+ | - true
+ | ...
+
+box.cfg{replication_synchro_quorum=quorum, 
replication_synchro_timeout=timeout}
+ | ---
+ | ...
+test_run:cmd('stop server replica')
+ | ---
+ | - true
+ | ...
+test_run:cmd('delete server replica')
+ | ---
+ | - true
+ | ...
+box.space.sync:drop()
+ | ---
+ | ...
+box.schema.user.revoke('guest', 'replication')
+ | ---
+ | ...
diff --git a/test/replication/sync_replication_sanity.test.lua 
b/test/replication/sync_replication_sanity.test.lua
index fd7ed537e..8715a4600 100644
--- a/test/replication/sync_replication_sanity.test.lua
+++ b/test/replication/sync_replication_sanity.test.lua
@@ -27,3 +27,55 @@ s2:drop()

  -- Local space can't be synchronous.
  box.schema.create_space('test', {is_sync = true, is_local = true})
+
+--
+-- gh-4847, gh-4848: CONFIRM and ROLLBACK entries in WAL.
+--
+env = require('test_run')
+test_run = env.new()
+fiber = require('fiber')
+engine = test_run:get_cfg('engine')
+
+box.schema.user.grant('guest', 'replication')
+-- Set up synchronous replication options.
+quorum = box.cfg.replication_synchro_quorum
+timeout = box.cfg.replication_synchro_timeout
+box.cfg{replication_synchro_quorum=2, replication_synchro_timeout=0.1}
+
+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')
+
+lsn = box.info.lsn
+box.space.sync:insert{1}
+-- 1 for insertion, 1 for CONFIRM message.
+box.info.lsn - lsn
+-- Raise quorum so that master has to issue a ROLLBACK.
+box.cfg{replication_synchro_quorum=3}
+t = fiber.time()
+box.space.sync:insert{2}
+-- Check that master waited for acks.
+fiber.time() - t > box.cfg.replication_synchro_timeout
+box.cfg{replication_synchro_quorum=2}
+box.space.sync:insert{3}
+box.space.sync:select{}
+
+-- Check consistency on replica.
+test_run:cmd('switch replica')
+box.space.sync:select{}
+
+-- Check consistency in recovered data.
+test_run:cmd('restart server replica')
+box.space.sync:select{}
+
+-- Cleanup.
+test_run:cmd('switch default')
+
+box.cfg{replication_synchro_quorum=quorum, 
replication_synchro_timeout=timeout}
+test_run:cmd('stop server replica')
+test_run:cmd('delete server replica')
+box.space.sync:drop()
+box.schema.user.revoke('guest', 'replication')

--

Serge Petrenko