File: rpl_parallel_recovery.test

package info (click to toggle)
mysql-8.0 8.0.43-3
  • links: PTS, VCS
  • area: main
  • in suites: sid
  • size: 1,273,924 kB
  • sloc: cpp: 4,684,605; ansic: 412,450; pascal: 108,398; java: 83,641; perl: 30,221; cs: 27,067; sql: 26,594; sh: 24,181; python: 21,816; yacc: 17,169; php: 11,522; xml: 7,388; javascript: 7,076; makefile: 2,194; lex: 1,075; awk: 670; asm: 520; objc: 183; ruby: 97; lisp: 86
file content (29 lines) | stat: -rw-r--r-- 1,185 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
#
# WL#5569/7165 MTS
#
# The test verifies MTS recovery in cases SQL applier threads (workers)
# have not completed all of their assignments at time they've been stopped.
# The state of the slave server might become insonsistent with that of the master
# 'cos there could be a number of event groups (transactions) uncommitted (gaps) followed
# with that of committed.
# The test proves that once the stop condition (e.g slave local error) is removed
# replication restarts, the gaps are filled up, and the eventual slave server state
# is consistent.
# The test is MTS Logical_clock specific due to a mechanism of failure injection.

# --let $rpl_topology= 1->2
# --let $rpl_debug= 1
# --source include/rpl_init.inc
# --source include/rpl_default_connections.inc

# Row format is chosen to ease the logged transaction size control
--source include/have_binlog_format_row.inc
--source include/force_myisam_default.inc
--source include/have_myisam.inc
--source include/not_group_replication_plugin.inc
--source include/have_replica_parallel_type_logical_clock.inc
--source include/master-slave.inc

--source extra/rpl_tests/rpl_parallel_recovery_core.test

--source include/rpl_end.inc