Uploaded image for project: 'MariaDB Server'
  1. MariaDB Server
  2. MDEV-4277

Crash inside mi_killed_in_mariadb() with myisammrg

Details

    • Bug
    • Status: Closed (View Workflow)
    • Major
    • Resolution: Fixed
    • 10.0.1, 5.5.30, 5.3.12
    • 10.0.2, 5.5.31, 5.3.13
    • None
    • None
    • RedHat EL 5.3 2.6.18-274.7.1.el5 x86_64 using bintar mariadb-5.3.12-Linux-x86_64.tar.gz

    Description

      Hello and thank you for mariadb-5.3.12-Linux-x86_64.tar.gz

      Had a crash on a recently upgraded (from MySQL 5.1.48) today, wondering if the backtrace might look like a known issue. I do not have a reproducer at the moment. Thanks.

       
      130315  9:43:03 [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 http://kb.askmonty.org/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: 5.3.12-MariaDB-log
      key_buffer_size=34359738368
      read_buffer_size=8388608
      max_used_connections=135
      max_threads=501
      thread_count=20
      connection_count=20
      It is possible that mysqld could use up to
      key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 40230290 K  bytes of memory
      Hope that's ok; if not, decrease some variables in the equation.
       
      Thread pointer: 0x0x2abbac0c2a10
      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 = 0x43e0f0f8 thread_stack 0x48000
      ./bin/mysqld(my_print_stacktrace+0x2e) [0x998a1e]
      ./bin/mysqld(handle_fatal_signal+0x3f9) [0x71ede9]
      /lib64/libpthread.so.0 [0x399f40eb70]
      ./bin/mysqld(mi_killed_in_mariadb+0xb) [0x7b412b]
      ./bin/mysqld(mi_yield_and_check_if_killed+0x23) [0x7caba3]
      ./bin/mysqld(mi_rnext_same+0xdb) [0x7e19fb]
      ./bin/mysqld(myrg_rnext_same+0x33) [0x86eb33]
      ./bin/mysqld(ha_myisammrg::index_next_same(unsigned char*, unsigned char const*, unsigned int)+0x20) [0x86cd60]
      ./bin/mysqld(handler::read_range_next()+0x3f) [0x70fc9f]
      ./bin/mysqld(handler::multi_range_read_next(void**)+0xd2) [0x7a52a2]
      ./bin/mysqld(QUICK_RANGE_SELECT::get_next()+0x39) [0x6ecad9]
      ./bin/mysqld(QUICK_ROR_INTERSECT_SELECT::get_next()+0x3c) [0x6ededc]
      ./bin/mysqld [0x70ac7e]
      ./bin/mysqld(sub_select(JOIN*, st_join_table*, bool)+0xae) [0x678a1e]
      ./bin/mysqld [0x678f54]
      ./bin/mysqld(JOIN::exec()+0x856) [0x68afb6]
      ./bin/mysqld(mysql_select(THD*, Item***, TABLE_LIST*, unsigned int, List<Item>&, Item*, unsigned int, st_order*, st_order*, Item*, st_order*, unsigned long long, select_result*, st_select_lex_unit*, st_select_lex*)+0x199) [0x68d0f9]
      ./bin/mysqld(handle_select(THD*, st_lex*, select_result*, unsigned long)+0x16f) [0x68dacf]
      ./bin/mysqld [0x601e9e]
      ./bin/mysqld(mysql_execute_command(THD*)+0x3a43) [0x607833]
      ./bin/mysqld(mysql_parse(THD*, char*, unsigned int, char const**)+0x299) [0x60a629]
      ./bin/mysqld(dispatch_command(enum_server_command, THD*, char*, unsigned int)+0xdf9) [0x60b879]
      ./bin/mysqld(do_command(THD*)+0x101) [0x60c0b1]
      ./bin/mysqld(handle_one_connection+0xfd) [0x5fd7bd]
      /lib64/libpthread.so.0 [0x399f40673d]
      /lib64/libc.so.6(clone+0x6d) [0x399e8d44bd]
       
      Trying to get some variables.
      Some pointers may be invalid and cause the dump to abort.
      Query (0x2abbc81173e8): select l.dttm as Date, l.usr as Username, l.serial as SN, l.lntype as Type, l.lntxt as Text from logh l  INNER JOIN session s  ON s.sessionid = l.sessionid where l.lntype in ('XXXXXXXXXXXXXXX')  and l.dt in ('2012-02-20')  and l.usr in ('YYYYYYYYYYY')  and l.serial in ('1234567')  and l.lntxt like '%Some text%' and s.speed != 'some_value' and s.inhouse = 'n' limit 1000
      Connection ID (thread ID): 3806879
      Status: NOT_KILLED
      Optimizer switch: index_merge=on,index_merge_union=on,index_merge_sort_union=on,index_merge_intersection=on,index_merge_sort_intersection=off,index_conditio                                n_pushdown=on,derived_merge=on,derived_with_keys=on,firstmatch=on,loosescan=on,materialization=on,in_to_exists=on,semijoin=on,partial_match_rowid_merge=on,p                                artial_match_table_scan=on,subquery_cache=on,mrr=off,mrr_cost_based=off,mrr_sort_keys=off,outer_join_with_cache=on,semijoin_with_cache=on,join_cache_increme                                ntal=on,join_cache_hashed=on,join_cache_bka=on,optimize_join_buffer_size=off,table_elimination=on
       
      The manual page at http://dev.mysql.com/doc/mysql/en/crashing.html contains
      information that should help you find out what is causing the crash.

      Attachments

        Activity

          elenst Elena Stepanova added a comment - - edited

          Hi Peter,

          I don't think we've seen any similar crashes recently.
          I presume you can't provide the database dump? Can you quote table structures for `logh` and `session` and underlying tables, and give a hint on how many rows they might have (approximately)? I'll try to generate similar load and see how it goes.

          elenst Elena Stepanova added a comment - - edited Hi Peter, I don't think we've seen any similar crashes recently. I presume you can't provide the database dump? Can you quote table structures for `logh` and `session` and underlying tables, and give a hint on how many rows they might have (approximately)? I'll try to generate similar load and see how it goes.

          Peter, thank you, I'm able to reproduce it based on the info you provided. Will add a test case soon.

          elenst Elena Stepanova added a comment - Peter, thank you, I'm able to reproduce it based on the info you provided. Will add a test case soon.

          That is great news. I am impressed. Thank you. I was contemplating a roll back, and may still have to do so if it repeats, so I am particularly interested if there is an optimizer switch I can turn off to avoid hitting the bug (or not).

          thatsafunnyname Peter (Stig) Edwards added a comment - That is great news. I am impressed. Thank you. I was contemplating a roll back, and may still have to do so if it repeats, so I am particularly interested if there is an optimizer switch I can turn off to avoid hitting the bug (or not).
          elenst Elena Stepanova added a comment - - edited

          Yes, in my version of the test case the crash happens with index_merge=on,index_merge_intersection=on. Turning off at least one of these prevents the crash (and changes the plan so there is no index_merge/intersect anymore).

          Of course it still might be different on your data, but I think it's worth trying.

          elenst Elena Stepanova added a comment - - edited Yes, in my version of the test case the crash happens with index_merge=on,index_merge_intersection=on. Turning off at least one of these prevents the crash (and changes the plan so there is no index_merge/intersect anymore). Of course it still might be different on your data, but I think it's worth trying.

          The test case is attached (mdev-4277.test). It's plain SQL which can be used both in MTR or in MySQL client.
          Minimal optimizer_switch: index_merge=on,index_merge_intersection=on

          elenst Elena Stepanova added a comment - The test case is attached (mdev-4277.test). It's plain SQL which can be used both in MTR or in MySQL client. Minimal optimizer_switch: index_merge=on,index_merge_intersection=on

          Committed a patch. Need a review discussion for it.

          psergei Sergei Petrunia added a comment - Committed a patch. Need a review discussion for it.

          The fix is pushed into MariaDB-5.3 tree.

          psergei Sergei Petrunia added a comment - The fix is pushed into MariaDB-5.3 tree.

          Thank you.

          thatsafunnyname Peter (Stig) Edwards added a comment - Thank you.

          People

            psergei Sergei Petrunia
            thatsafunnyname Peter (Stig) Edwards
            Votes:
            0 Vote for this issue
            Watchers:
            4 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.