Uploaded image for project: 'MariaDB Server'
  1. MariaDB Server
  2. MDEV-31838

Assertion fails in binlog_log_row_online_alter upon parallel replication with two-phase alter and MINIMAL binlog_row_image

Details

    Description

      I don't know why IGNORE is important in the update. It causes concerns in itself, as there is nothing to ignore, UPDATE is valid.

      Run the test case with --mysqld=--binlog-alter-two-phase=ON --mysqld=--slave_parallel_threads=4 --mysqld=--binlog_row_image=MINIMAL
      Somehow setting them dynamically doesn't cause the same effect, or maybe I'm not doing it correctly.

      --source include/have_debug_sync.inc
      		 
      --source include/have_binlog_format_row.inc
      		 
      --source include/master-slave.inc
      		 
       
      		 
      CREATE TABLE t (id INT, a INT, PRIMARY KEY(id)) ENGINE=Aria;
      		 
      INSERT INTO t VALUES (1,0);
      		 
       
      		 
      --connection master1
      		 
      --send set debug_sync= 'now wait_for go_dml'
      		 
       
      		 
      --connection master
      		 
      set debug_sync= 'alter_table_online_progress signal go_dml wait_for go_alter';
      		 
      --send
      		 
      ALTER TABLE t FORCE, ALGORITHM=COPY, LOCK=NONE;
      		 
       
      		 
      --connection master1
      		 
      --reap
      		 
      UPDATE IGNORE t SET a = 1;
      		 
      set debug_sync= 'now signal go_alter';
      		 
       
      		 
      --connection master
      		 
      --reap
      		 
       
      		 
      --sync_slave_with_master
      		 
       
      		 
      --connection master
      		 
      DROP TABLE t;
      		 
      --source include/rpl_end.inc

      bb-11.2-oalter 67d0139c

      		 
      mariadbd: /data/src/bb-11.2-oalter-debug/sql/log.cc:2285: int binlog_log_row_online_alter(TABLE*, const uchar*, const uchar*, bool (*)(THD*, TABLE*, Event_log*, binlog_cache_data*, bool, ulong, const uchar*, const uchar*)): Assertion `!before_record || bitmap_is_set_all(table->read_set)' failed.
      		 
      230803 19:03:18 [ERROR] mysqld got signal 6 ;
      		 
       
      		 
      #9  0x00007f1d04453df2 in __GI___assert_fail (assertion=0x5591f232e1e0 "!before_record || bitmap_is_set_all(table->read_set)", file=0x5591f232d0a8 "/data/src/bb-11.2-oalter-debug/sql/log.cc", line=2285, function=0x5591f232e218 "int binlog_log_row_online_alter(TABLE*, const uchar*, const uchar*, bool (*)(THD*, TABLE*, Event_log*, binlog_cache_data*, bool, ulong, const uchar*, const uchar*))") at ./assert/assert.c:101
      		 
      #10 0x00005591f1838634 in binlog_log_row_online_alter (table=0x7f1c7c015ef8, before_record=0x7f1c7c016388 "\375\001", after_record=0x7f1c7c016378 "\375\001", log_func=0x5591f16f6454 <Update_rows_log_event::binlog_row_logging_function(THD*, TABLE*, Event_log*, binlog_cache_data*, bool, unsigned long, unsigned char const*, unsigned char const*)>) at /data/src/bb-11.2-oalter-debug/sql/log.cc:2285
      		 
      #11 0x00005591f16ee848 in handler::binlog_log_row (this=0x7f1c7c016800, before_record=0x7f1c7c016388 "\375\001", after_record=0x7f1c7c016378 "\375\001", log_func=0x5591f16f6454 <Update_rows_log_event::binlog_row_logging_function(THD*, TABLE*, Event_log*, binlog_cache_data*, bool, unsigned long, unsigned char const*, unsigned char const*)>) at /data/src/bb-11.2-oalter-debug/sql/handler.cc:7295
      		 
      #12 0x00005591f16f1663 in handler::ha_update_row (this=0x7f1c7c016800, old_data=0x7f1c7c016388 "\375\001", new_data=0x7f1c7c016378 "\375\001") at /data/src/bb-11.2-oalter-debug/sql/handler.cc:7889
      		 
      #13 0x00005591f1889b39 in Update_rows_log_event::do_exec_row (this=0x7f1c9003fb58, rgi=0x7f1c9003d950) at /data/src/bb-11.2-oalter-debug/sql/log_event_server.cc:8018
      		 
      #14 0x00005591f18801e6 in Rows_log_event::do_apply_event (this=0x7f1c9003fb58, rgi=0x7f1c9003d950) at /data/src/bb-11.2-oalter-debug/sql/log_event_server.cc:5131
      		 
      #15 0x00005591f186a36c in Log_event::apply_event (this=0x7f1c9003fb58, rgi=0x7f1c9003d950) at /data/src/bb-11.2-oalter-debug/sql/log_event.cc:3880
      		 
      #16 0x00005591f11af347 in apply_event_and_update_pos_apply (ev=0x7f1c9003fb58, thd=0x7f1c7c000dc8, rgi=0x7f1c9003d950, reason=0) at /data/src/bb-11.2-oalter-debug/sql/slave.cc:3877
      		 
      #17 0x00005591f11afbba in apply_event_and_update_pos_for_parallel (ev=0x7f1c9003fb58, thd=0x7f1c7c000dc8, rgi=0x7f1c9003d950) at /data/src/bb-11.2-oalter-debug/sql/slave.cc:4074
      		 
      #18 0x00005591f15874d4 in rpt_handle_event (qev=0x7f1c9003fdf8, rpt=0x7f1c9000d420) at /data/src/bb-11.2-oalter-debug/sql/rpl_parallel.cc:61
      		 
      #19 0x00005591f158b38b in handle_rpl_parallel_thread (arg=0x7f1c9000d420) at /data/src/bb-11.2-oalter-debug/sql/rpl_parallel.cc:1429
      		 
      #20 0x00005591f19fb438 in pfs_spawn_thread (arg=0x7f1c9000e8c8) at /data/src/bb-11.2-oalter-debug/storage/perfschema/pfs.cc:2201
      		 
      #21 0x00007f1d044a7fd4 in start_thread (arg=<optimized out>) at ./nptl/pthread_create.c:442
      		 
      #22 0x00007f1d045285bc in clone3 () at ../sysdeps/unix/sysv/linux/x86_64/clone3.S:81

      Attachments

        Issue Links

          is caused by

          Task - A task that needs to be done. MDEV-16329 Engine-independent online ALTER TABLE

          • Blocker - Blocks development and/or testing work, production could not run.
          • Closed

          Activity

            Ascending order - Click to sort in descending order
            elenst Elena Stepanova created issue -
            elenst Elena Stepanova made changes -
            Field Original Value New Value
            nikitamalyavin Nikita Malyavin made changes -
            Status Open [ 1 ] In Progress [ 3 ]
            nikitamalyavin Nikita Malyavin made changes -
            Assignee Nikita Malyavin [ nikitamalyavin ] Sergei Golubchik [ serg ]
            Status In Progress [ 3 ] In Review [ 10002 ]
            serg Sergei Golubchik added a comment -

            looks ok, but change the commit comment, please.
            something like

            when a slave thread is applying replicated row events in parallel with the running online alter, we need all columns to be read from the table (for the online alter logged row event) not only those that were present in the pre-image for the replicated row event.

            serg Sergei Golubchik added a comment - looks ok, but change the commit comment, please. something like when a slave thread is applying replicated row events in parallel with the running online alter, we need all columns to be read from the table (for the online alter logged row event) not only those that were present in the pre-image for the replicated row event.
            serg Sergei Golubchik made changes -
            Assignee Sergei Golubchik [ serg ] Nikita Malyavin [ nikitamalyavin ]
            Status In Review [ 10002 ] Stalled [ 10000 ]
            nikitamalyavin Nikita Malyavin made changes -
            Status Stalled [ 10000 ] In Testing [ 10301 ]
            serg Sergei Golubchik made changes -
            Fix Version/s 11.2.1 [ 29034 ]
            Fix Version/s 11.2 [ 28603 ]
            Resolution Fixed [ 1 ]
            Status In Testing [ 10301 ] Closed [ 6 ]

            People

              nikitamalyavin Nikita Malyavin
              elenst Elena Stepanova
              Votes:
              0 Vote for this issue
              Watchers:
              2 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.