Uploaded image for project: 'MariaDB Server'
  1. MariaDB Server
  2. MDEV-21473

Galera node that is also an async slave from other cluster may crash unexpectedly on rollback or commit

    XMLWordPrintable

Details

    Description

      Due to a bug in interaction between native slave thread processing and MariaDB wsrep patch we may end up with crashes like these:

      ...
      2020-01-09 11:16:38 140078043080448 [Note] WSREP: Member 2.0 (hss6geo246o) synced with group.
      2020-01-09 12:18:41 140076876134144 [ERROR] Master 'to_master_0': Slave SQL: Node has dropped from cluster, Gtid 200-200-3694, Internal MariaDB error code: 1047
      2020-01-09 12:18:41 140076876134144 [Note] Master 'to_master_0': Slave SQL thread exiting, replication stopped in log 'binlog.000004' at position 10915199; GTID position '100-100-1123,200-200-3693'
      2020-01-09 12:18:41 140076876134144 [Note] Master 'to_master_0': WSREP: Slave error due to node temporarily non-primarySQL slave will continue
      2020-01-09 12:18:41 140076876134144 [Note] Master 'to_master_0': WSREP: slave restart: 3
      2020-01-09 12:18:41 140076876134144 [Note] Master 'to_master_0': WSREP: ready state reached
      ...
      2020-01-09 12:18:41 140076876134144 [Note] Master 'to_master_0': WSREP: Slave error due to node temporarily non-primarySQL slave will continue
      2020-01-09 12:18:41 140076876134144 [Note] Master 'to_master_0': WSREP: slave restart: 3
      2020-01-09 12:18:41 140076876134144 [Note] Master 'to_master_0': WSREP: ready state reached
      2020-01-09 12:18:41 140076876134144 [Note] Master 'to_master_0': Slave SQL thread initialized, starting replication in log 'binlog.000004' at position 10991571, relay log '/var/lib/mybinlog/relaylog-to_master_0.000006' position: 10407768; GTID position '100-100-1123,200-200-3834'
      2020-01-09 12:18:41 140078447146752 [ERROR] Master 'to_master_0': WSREP: FSM: no such a transition REPLICATING -> ROLLED_BACK
      200109 12:18:41 [ERROR] mysqld got signal 6 ;
      This could be because you hit a bug. It is also possible that this binary
      or one of the libraries it was linked against is corrupt, improperly built,
      or misconfigured. This error can also be caused by malfunctioning hardware.
       
      To report this bug, see https://mariadb.com/kb/en/reporting-bugs
       
      We will try our best to scrape up some info that will hopefully help
      diagnose the problem, but since we have already crashed, 
      something is definitely wrong and this may fail.
       
      Server version: 10.2.30-MariaDB-log
      key_buffer_size=134217728
      read_buffer_size=131072
      max_used_connections=41
      max_threads=153
      thread_count=86
      It is possible that mysqld could use up to 
      key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 467243 K  bytes of memory
      Hope that's ok; if not, decrease some variables in the equation.
       
      Thread pointer: 0x7f66686d5cd8
      Attempting backtrace. You can use the following information to find out
      where mysqld died. If you see no messages after this, something went
      terribly wrong...
      stack_bottom = 0x7f66307161a0 thread_stack 0x49000
      /usr/sbin/mysqld(my_print_stacktrace+0x2e)[0x55e16de2b47e]
      /usr/sbin/mysqld(handle_fatal_signal+0x30d)[0x55e16d8b0abd]
      sigaction.c:0(__restore_rt)[0x7f668de916d0]
      :0(__GI_raise)[0x7f668c394277]
      :0(__GI_abort)[0x7f668c395968]
      src/fsm.hpp:104(galera::FSM<galera::TrxHandle::State, galera::TrxHandle::Transition, galera::EmptyGuard, galera::EmptyAction>::shift_to(galera::TrxHandle::State))[0x7f6684fa61dc]
      :0(galera::ReplicatorSMM::post_rollback(galera::TrxHandle*))[0x7f6684f9c6a6]
      src/gu_mutex.hpp:38(gu::Mutex::unlock() const)[0x7f6684fad385]
      /usr/sbin/mysqld(+0x6148b5)[0x55e16d8408b5]
      /usr/sbin/mysqld(+0x6149e8)[0x55e16d8409e8]
      /usr/sbin/mysqld(_Z17ha_rollback_transP3THDb+0xe3)[0x55e16d8b3a73]
      /usr/sbin/mysqld(_Z14trans_rollbackP3THD+0x49)[0x55e16d7f80f9]
      /usr/sbin/mysqld(_Z21wsrep_client_rollbackP3THD+0x60)[0x55e16d856810]
      /usr/sbin/mysqld(+0x62acab)[0x55e16d856cab]
      /usr/sbin/mysqld(start_wsrep_THD+0x4fb)[0x55e16d84575b]
      pthread_create.c:0(start_thread)[0x7f668de89e25]
      /lib64/libc.so.6(clone+0x6d)[0x7f668c45cbad]
       
      Trying to get some variables.
      Some pointers may be invalid and cause the dump to abort.
      Query (0x0): 
      Connection ID (thread ID): 2571
      Status: NOT_KILLED
      ...
      

      or:

      ...
      2020-01-09 12:41:56 139923526936320 [Note] Master 'to_master_0': Slave SQL thread exiting, replication stopped in log 'binlog.000004' at position 76754168; GTID position '100-100-29316,200-200-123484'
      2020-01-09 12:41:56 139923526936320 [Note] Master 'to_master_0': WSREP: Slave error due to node temporarily non-primarySQL slave will continue
      2020-01-09 12:41:56 139923526936320 [Note] Master 'to_master_0': WSREP: slave restart: 3
      2020-01-09 12:41:56 139923526936320 [Note] Master 'to_master_0': WSREP: ready state reached
      2020-01-09 12:41:56 139923526936320 [Note] Master 'to_master_0': Slave SQL thread initialized, starting replication in log 'binlog.000004' at position 76754168, relay log '/var/lib/mybinlog/relaylog-to_master_0.000002' position: 17291006; GTID position '100-100-29316,200-200-123484'
      200109 12:41:56 [ERROR] mysqld got signal 11 ;
      This could be because you hit a bug. It is also possible that this binary
      or one of the libraries it was linked against is corrupt, improperly built,
      or misconfigured. This error can also be caused by malfunctioning hardware.
       
      To report this bug, see https://mariadb.com/kb/en/reporting-bugs
       
      We will try our best to scrape up some info that will hopefully help
      diagnose the problem, but since we have already crashed, 
      something is definitely wrong and this may fail.
       
      Server version: 10.2.30-MariaDB-log
      key_buffer_size=134217728
      read_buffer_size=131072
      max_used_connections=23
      max_threads=153
      thread_count=55
      It is possible that mysqld could use up to 
      key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 467243 K  bytes of memory
      Hope that's ok; if not, decrease some variables in the equation.
       
      Thread pointer: 0x7f4260316ee8
      Attempting backtrace. You can use the following information to find out
      where mysqld died. If you see no messages after this, something went
      terribly wrong...
      stack_bottom = 0x7f427c1e51a0 thread_stack 0x49000
      /usr/sbin/mysqld(my_print_stacktrace+0x2e)[0x562911c4c47e]
      /usr/sbin/mysqld(handle_fatal_signal+0x30d)[0x5629116d1abd]
      sigaction.c:0(__restore_rt)[0x7f42cd70b6d0]
      /usr/sbin/mysqld(_Z19wsrep_register_htonP3THDb+0x84)[0x562911661a84]
      /usr/sbin/mysqld(_Z12trans_commitP3THD+0x10d)[0x562911618f6d]
      /usr/sbin/mysqld(_ZN13Xid_log_event14do_apply_eventEP14rpl_group_info+0xca)[0x5629117c07ba]
      /usr/sbin/mysqld(+0x46d9eb)[0x5629114ba9eb]
      /usr/sbin/mysqld(handle_slave_sql+0x1533)[0x5629114c38b3]
      pthread_create.c:0(start_thread)[0x7f42cd703e25]
      /lib64/libc.so.6(clone+0x6d)[0x7f42cbcd6bad]
       
      Trying to get some variables.
      Some pointers may be invalid and cause the dump to abort.
      Query (0x0): 
      Connection ID (thread ID): 1103
      Status: NOT_KILLED
      

      or even like this:

      2020-01-09 12:18:41 140076876134144 [Note] Master 'to_master_0': WSREP: slave restart: 3
      2020-01-09 12:18:41 140076876134144 [Note] Master 'to_master_0': WSREP: ready state reached
      2020-01-09 12:18:41 140076876134144 [Note] Master 'to_master_0': Slave SQL thread initialized, starting replication in log 'binlog.000004' at position 10991571, relay log '/var/lib/mybinlog/relaylog-to_master_0.000006' position: 10407768; GTID position '100-100-1123,200-200-3834'
      2020-01-09 12:18:41 140078447146752 [ERROR] Master 'to_master_0': WSREP: FSM: no such a transition REPLICATING -> ROLLED_BACK
      200109 12:18:41 [ERROR] mysqld got signal 6 ;
      This could be because you hit a bug. It is also possible that this binary
      or one of the libraries it was linked against is corrupt, improperly built,
      or misconfigured. This error can also be caused by malfunctioning hardware.
       
      To report this bug, see https://mariadb.com/kb/en/reporting-bugs
       
      We will try our best to scrape up some info that will hopefully help
      diagnose the problem, but since we have already crashed, 
      something is definitely wrong and this may fail.
       
      Server version: 10.2.30-MariaDB-log
      key_buffer_size=134217728
      read_buffer_size=131072
      max_used_connections=41
      max_threads=153
      thread_count=86
      It is possible that mysqld could use up to 
      key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 467243 K  bytes of memory
      Hope that's ok; if not, decrease some variables in the equation.
       
      Thread pointer: 0x7f66686d5cd8
      Attempting backtrace. You can use the following information to find out
      where mysqld died. If you see no messages after this, something went
      terribly wrong...
      stack_bottom = 0x7f66307161a0 thread_stack 0x49000
      /usr/sbin/mysqld(my_print_stacktrace+0x2e)[0x55e16de2b47e]
      /usr/sbin/mysqld(handle_fatal_signal+0x30d)[0x55e16d8b0abd]
      sigaction.c:0(__restore_rt)[0x7f668de916d0]
      :0(__GI_raise)[0x7f668c394277]
      :0(__GI_abort)[0x7f668c395968]
      src/fsm.hpp:104(galera::FSM<galera::TrxHandle::State, galera::TrxHandle::Transition, galera::EmptyGuard, galera::EmptyAction>::shift_to(galera::TrxHandle::State))[0x7f6684fa61dc]
      :0(galera::ReplicatorSMM::post_rollback(galera::TrxHandle*))[0x7f6684f9c6a6]
      src/gu_mutex.hpp:38(gu::Mutex::unlock() const)[0x7f6684fad385]
      /usr/sbin/mysqld(+0x6148b5)[0x55e16d8408b5]
      /usr/sbin/mysqld(+0x6149e8)[0x55e16d8409e8]
      /usr/sbin/mysqld(_Z17ha_rollback_transP3THDb+0xe3)[0x55e16d8b3a73]
      /usr/sbin/mysqld(_Z14trans_rollbackP3THD+0x49)[0x55e16d7f80f9]
      /usr/sbin/mysqld(_Z21wsrep_client_rollbackP3THD+0x60)[0x55e16d856810]
      /usr/sbin/mysqld(+0x62acab)[0x55e16d856cab]
      /usr/sbin/mysqld(start_wsrep_THD+0x4fb)[0x55e16d84575b]
      pthread_create.c:0(start_thread)[0x7f668de89e25]
      /lib64/libc.so.6(clone+0x6d)[0x7f668c45cbad]
       
      Trying to get some variables.
      Some pointers may be invalid and cause the dump to abort.
      Query (0x0): 
      Connection ID (thread ID): 2571
      Status: NOT_KILLED
      

      Attachments

        Issue Links

          Activity

            People

              jplindst Jan Lindström (Inactive)
              valerii Valerii Kravchuk
              Votes:
              0 Vote for this issue
              Watchers:
              1 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved:

                Git Integration

                  Error rendering 'com.xiplink.jira.git.jira_git_plugin:git-issue-webpanel'. Please contact your Jira administrators.