Uploaded image for project: 'MariaDB Server'
  1. MariaDB Server
  2. MDEV-22330

mysqlbinlog stops with an error Don't know how to handle column type: 255 meta: 4 (0004)

Details

    Description

      From some time now when I try to use mysqlbinlog with the --verbose option I get an error:
      Error: Don't know how to handle column type: 255 meta: 4 (0004)

      If I don't use the --verbose option it seams to work fine but I cannot see the query executed.

      Since the binlog file is too large I'm attaching here only the output that I've narrowed using --start-datetime --stop-datetime from the command without the verbose option.

      If you need the binlog file itself, please give instruction how to send it in a secure way since I don't want the hole file to be made available public.

      Attachments

        Activity

          lukav Anton Avramov added a comment -

          The problem seams to be with a column of type geometry DEFAULT NULL
          And the insert statement has a value of null for that column

          lukav Anton Avramov added a comment - The problem seams to be with a column of type geometry DEFAULT NULL And the insert statement has a value of null for that column

          Thanks for the report and for narrowing down the problem.

          Reproducible as described on 10.3-10.5, including old 10.3 releases.
          Not reproducible on 10.2.

          --source include/have_binlog_format_row.inc
           
          RESET MASTER;
          create table t1 (a GEOMETRY DEFAULT NULL);
          insert into t1 values (NULL);
          FLUSH BINARY LOGS;
          --let $datadir= `select @@datadir`
          --exec $MYSQL_BINLOG --verbose $datadir/master-bin.000001
           
          # Cleanup
          DROP TABLE t1;
          

          10.3 644d9f38

          RESET MASTER;
          create table t1 (a GEOMETRY DEFAULT NULL);
          insert into t1 values (NULL);
          FLUSH BINARY LOGS;
          bug.t                                    [ fail ]
                  Test ended at 2020-05-05 02:13:22
           
          CURRENT_TEST: bug.t
           
          Error: Don't know how to handle column type: 255 meta: 4 (0004)
          mysqltest: At line 8: exec of '/data/bld/10.3-debug/bin/mysqlbinlog --defaults-file=/data/bld/10.3-debug/mysql-test/var/my.cnf --local-load=/data/bld/10.3-debug/mysql-test/var/tmp --verbose /data/bld/10.3-debug/mysql-test/var/mysqld.1/data//master-bin.000001' failed, error: 256, status: 1, errno: 11
          Output from before failure:
          

          elenst Elena Stepanova added a comment - Thanks for the report and for narrowing down the problem. Reproducible as described on 10.3-10.5, including old 10.3 releases. Not reproducible on 10.2. --source include/have_binlog_format_row.inc   RESET MASTER; create table t1 (a GEOMETRY DEFAULT NULL ); insert into t1 values ( NULL ); FLUSH BINARY LOGS; --let $datadir= `select @@datadir` --exec $MYSQL_BINLOG --verbose $datadir/master-bin.000001   # Cleanup DROP TABLE t1; 10.3 644d9f38 RESET MASTER; create table t1 (a GEOMETRY DEFAULT NULL); insert into t1 values (NULL); FLUSH BINARY LOGS; bug.t [ fail ] Test ended at 2020-05-05 02:13:22   CURRENT_TEST: bug.t   Error: Don't know how to handle column type: 255 meta: 4 (0004) mysqltest: At line 8: exec of '/data/bld/10.3-debug/bin/mysqlbinlog --defaults-file=/data/bld/10.3-debug/mysql-test/var/my.cnf --local-load=/data/bld/10.3-debug/mysql-test/var/tmp --verbose /data/bld/10.3-debug/mysql-test/var/mysqld.1/data//master-bin.000001' failed, error: 256, status: 1, errno: 11 Output from before failure:

          The issue exists in 10.1 and above. The error message gets printed as part of comment as shown below

          #200811 16:43:38 server id 1  end_log_pos 566 	Write_rows: table id 31 flags: STMT_END_F
           
          BINLOG '
          Yn0yXxMBAAAAKgAAABgCAAAAAB8AAAAAAAEABHRlc3QAAnQxAAH/AQQB
          Yn0yXxcBAAAAHgAAADYCAAAAAB8AAAAAAAEAAf//
          ### INSERT INTO `test`.`t1`
          ### SET
          ###   @1=!! Don't know how to handle column type=255 meta=4 (0004)'/*!*/;
          # at 566
          #200811 16:43:38 server id 1  end_log_pos 635 	Query	thread_id=4	exec_time=0	error_code=0
          SET TIMESTAMP=1597144418/*!*/;
          COMMIT
          

          sujatha.sivakumar Sujatha Sivakumar (Inactive) added a comment - The issue exists in 10.1 and above. The error message gets printed as part of comment as shown below #200811 16:43:38 server id 1 end_log_pos 566 Write_rows: table id 31 flags: STMT_END_F   BINLOG ' Yn0yXxMBAAAAKgAAABgCAAAAAB8AAAAAAAEABHRlc3QAAnQxAAH/AQQB Yn0yXxcBAAAAHgAAADYCAAAAAB8AAAAAAAEAAf// ### INSERT INTO `test`.`t1` ### SET ### @1=!! Don't know how to handle column type=255 meta=4 (0004)'/*!*/; # at 566 #200811 16:43:38 server id 1 end_log_pos 635 Query thread_id=4 exec_time=0 error_code=0 SET TIMESTAMP=1597144418/*!*/; COMMIT
          sujatha.sivakumar Sujatha Sivakumar (Inactive) added a comment - Hello holyfoot Please review the fix for MDEV-22330 . Patch: https://github.com/MariaDB/server/commit/5dc2ebae47481afe5b0769f0e0463d64b77143c0 Test results: http://buildbot.askmonty.org/buildbot/grid?category=main&branch=bb-10.1-sujatha Thank you.

          Hello Andrei,

          Can you please review the following changes.
          Since the patch is specific to GEOMETRY type I requested Holyfoot's review and it is complete.

          Patch: https://github.com/MariaDB/server/commit/5dc2ebae47481afe5b0769f0e0463d64b77143c0

          Test results: http://buildbot.askmonty.org/buildbot/grid?category=main&branch=bb-10.1-sujatha

          sujatha.sivakumar Sujatha Sivakumar (Inactive) added a comment - Hello Andrei, Can you please review the following changes. Since the patch is specific to GEOMETRY type I requested Holyfoot's review and it is complete. Patch: https://github.com/MariaDB/server/commit/5dc2ebae47481afe5b0769f0e0463d64b77143c0 Test results: http://buildbot.askmonty.org/buildbot/grid?category=main&branch=bb-10.1-sujatha
          Elkin Andrei Elkin added a comment -

          The patch looks good but I am not sure about the produced comments' usability. Please clarify.

          Elkin Andrei Elkin added a comment - The patch looks good but I am not sure about the produced comments' usability. Please clarify.

          Fix pushed to 10.1.47. Merged changes upto 10.4.

          sujatha.sivakumar Sujatha Sivakumar (Inactive) added a comment - Fix pushed to 10.1.47. Merged changes upto 10.4.

          People

            sujatha.sivakumar Sujatha Sivakumar (Inactive)
            lukav Anton Avramov
            Votes:
            0 Vote for this issue
            Watchers:
            5 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.