Angelique Sklavounos (Inactive) added a comment - 2023-04-25 15:26 Looking at the tests that run before these failing tests: rpl.rpl_ssl 'mix' w1 [ skipped ] Need "--big-test" when running with Valgrind binlog.binlog_server_id 'stmt' w34 [ pass ] 115 binlog.binlog_stm_ctype_cp932 'mix' w16 [ pass ] 13276 binlog.binlog_stm_drop_tbl 'mix' w16 [ pass ] 148 binlog.binlog_spurious_ddl_errors 'innodb,stmt' w40 [ pass ] 2447 rpl.rpl_ps 'row' w24 [ pass ] 1308 rpl.rpl_ssl1 'mix' w1 [ fail ] rpl.rpl_mdev12179 'innodb,mix' w4 [ skipped ] Need "--big-test" when running with Valgrind sys_vars.explicit_defaults_for_timestamp_on w6 [ pass ] 908 oqgraph.general-innodb 'innodb' w12 [ pass ] 14198 oqgraph.general-MyISAM w30 [ pass ] 3327 versioning.replace 'heap,innodb,pk' w33 [ pass ] 214 main.log_state_bug33693 w37 [ pass ] 11 rpl.rpl_temporary 'mix' w40 [ pass ] 4474 rpl.rpl_set_statement_default_master 'row' w8 [ pass ] 923 rpl.rpl_gtid_until 'innodb,stmt' w11 [ pass ] 11348 versioning.replace 'heap,innodb,sec' w33 [ skipped ] pk or unique only rpl.rpl_row_trig001 'row' w34 [ pass ] 1818 oqgraph.invalid_operations w30 [ pass ] 123 rpl.rpl_multi_engine 'innodb,mix' w4 [ fail ] Both rpl.rpl_ssl and rpl.rpl_mdev12179 call the macro no_valgrind_without_big.inc AFTER rpl_init.inc is called. So, the slave server is set up with START SLAVE , but then the test is skipped, yet the slave is never stopped: CURRENT_TEST: rpl.rpl_ssl 2023-04-24 10:47:39 265 [Note] Deleted Master_info file '/buildbot/amd64-ubuntu-2204-valgrind/build/mysql-test/var/1/mysqld.2/data/master.info'. 2023-04-24 10:47:39 265 [Note] Deleted Master_info file '/buildbot/amd64-ubuntu-2204-valgrind/build/mysql-test/var/1/mysqld.2/data/relay-log.info'. 2023-04-24 10:47:39 265 [Note] Master connection name: '' Master_info_file: 'master.info' Relay_info_file: 'relay-log.info' 2023-04-24 10:47:39 265 [Note] 'CHANGE MASTER TO executed'. Previous state master_host='127.0.0.1', master_port='16000', master_log_file='', master_log_pos='4'. New state master_host='127.0.0.1', master_port='16000', master_log_file='master-bin.000001', master_log_pos='4'. 2023-04-24 10:47:39 267 [Note] Slave I/O thread: Start asynchronous replication to master 'root@127.0.0.1:16000' in log 'master-bin.000001' at position 4 2023-04-24 10:47:39 268 [Note] Slave SQL thread initialized, starting replication in log 'master-bin.000001' at position 4, relay log './slave-relay-bin.000001' position: 4 2023-04-24 10:47:39 267 [Note] Slave I/O thread: connected to master 'root@127.0.0.1:16000',replication started in log 'master-bin.000001' at position 4 CURRENT_TEST: rpl.rpl_ssl1 2023-04-24 10:47:40 0 [Note] /buildbot/amd64-ubuntu-2204-valgrind/build/sql/mysqld (initiated by: unknown): Normal shutdown 2023-04-24 10:47:40 0 [Note] Event Scheduler: Purging the queue. 0 events 2023-04-24 10:47:40 268 [Note] Error reading relay log event: slave SQL thread was killed 2023-04-24 10:47:40 268 [Note] Slave SQL thread exiting, replication stopped in log 'master-bin.000001' at position 329 This can be reproduced locally with ./mtr rpl.rpl_ssl,mix rpl.rpl_ssl1,mix --no-reorder --valgrind (and a hack to the MTR script to disallow big tests to run automatically when test cases are specified). The fix is to call no_valgrind_without_big.inc before the master and slave servers are set up.

Angelique Sklavounos (Inactive) added a comment - 2023-04-28 16:53 The fix was introduced, along with re-ordering to do other macros that check test environment capabilities before master/slave is set up. In the test runs, rpl.rpl_ssl1 and rpl.rpl_multi_engine passed: https://buildbot.mariadb.org/#/builders/551/builds/1671/steps/7/logs/stdio https://buildbot.mariadb.org/#/builders/551/builds/1664/steps/7/logs/stdio