From: Rob Wultsch Date: December 15 2011 7:47am Subject: Re: 'myisam_use_mmap' unstable like hell List-Archive: http://lists.mysql.com/mysql/226480 Message-Id: MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable To be brutally honest, if you want stability you should not be using MyISAM, much less a not particularly commonly used feature. On Thu, Nov 24, 2011 at 12:58 AM, Reindl Harald wr= ote: > and the next one without "memlock" > > 111124 09:50:30 mysqld_safe mysqld from pid file /var/run/mysqld/mysqld.p= id ended > 111124 09:50:35 mysqld_safe Starting mysqld daemon with databases from /V= olumes/dune/mysql_data > 111124 =A09:50:35 [Note] Plugin 'InnoDB' is disabled. > 111124 =A09:50:35 [Note] Plugin 'FEDERATED' is disabled. > 111124 =A09:50:35 [Note] Plugin 'BLACKHOLE' is disabled. > 111124 =A09:50:35 [Note] Plugin 'ARCHIVE' is disabled. > 111124 =A09:50:35 [Note] Plugin 'partition' is disabled. > 111124 =A09:50:35 [Note] Event Scheduler: Loaded 0 events > 111124 =A09:50:35 [Note] /usr/libexec/mysqld: ready for connections. > Version: '5.5.18-log' =A0socket: '/var/lib/mysql/mysql.sock' =A0port: 330= 6 =A0thelounge.net build > 111124 =A09:53:12 [ERROR] Got error 127 when reading table './aume/skefon= ds2009_ext_content' > 111124 =A09:53:12 [ERROR] Got error 127 when reading table './aume/skefon= ds2009_ext_content' > 111124 =A09:53:17 [ERROR] Got error 127 when reading table './aume/skefon= ds2009_ext_content' > 111124 =A09:53:17 [ERROR] Got error 127 when reading table './aume/skefon= ds2009_ext_content' > 111124 =A09:53:22 [ERROR] Got error 127 when reading table './aume/skefon= ds2009_ext_content' > 111124 =A09:53:22 [ERROR] Got error 127 when reading table './aume/skefon= ds2009_ext_content' > 111124 =A09:53:32 [ERROR] Got error 127 when reading table './afi/cms1_su= b2' > 111124 =A09:53:32 [ERROR] Got error 127 when reading table './afi/cms1_su= b2' > 111124 =A09:55:02 [ERROR] Got error 127 when reading table './hurnaus/cms= 1_galerie_sub' > 111124 =A09:55:02 [ERROR] Got error 127 when reading table './hurnaus/cms= 1_galerie_sub' > 111124 =A09:55:14 - 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 buil= t, > or misconfigured. This error can also be caused by malfunctioning hardwar= e. > We will try our best to scrape up some info that will hopefully help diag= nose > the problem, but since we have already crashed, something is definitely w= rong > and this may fail. > > key_buffer_size=3D268435456 > read_buffer_size=3D262144 > max_used_connections=3D12 > max_threads=3D200 > thread_count=3D3 > connection_count=3D3 > It is possible that mysqld could use up to > key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads =3D 4= 18015 K > bytes of memory > Hope that's ok; if not, decrease some variables in the equation. > > Thread pointer: 0x2ea7080 > 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 =3D 0x7ffd2ea39d40 thread_stack 0x40000 > /usr/libexec/mysqld(my_print_stacktrace+0x33)[0x7ab8f3] > /usr/libexec/mysqld(handle_segfault+0x470)[0x50f190] > /lib64/libpthread.so.0(+0xeeb0)[0x7ffdaae93eb0] > /lib64/libc.so.6(+0x12ffa5)[0x7ffda920cfa5] > /usr/libexec/mysqld(mi_mmap_pread+0x15a)[0x90880a] > /usr/libexec/mysqld(_mi_read_dynamic_record+0x1fe)[0x90ac5e] > /usr/libexec/mysqld(mi_rkey+0x378)[0x930f48] > /usr/libexec/mysqld(_ZN9ha_myisam14index_read_mapEPhPKhm16ha_rkey_functio= n+0x59)[0x8f1fe9] > /usr/libexec/mysqld[0x5b3f35] > /usr/libexec/mysqld(_Z10sub_selectP4JOINP13st_join_tableb+0x61)[0x5a4721] > /usr/libexec/mysqld[0x5b2c65] > /usr/libexec/mysqld(_ZN4JOIN4execEv+0xbe1)[0x5c39b1] > /usr/libexec/mysqld(_Z12mysql_selectP3THDPPP4ItemP10TABLE_LISTjR4ListIS1_= ES2_jP8st_orderSB_S2_SB_yP13select_resultP18st_select_lex_unitP13st_select_= lex+0x152)[0x5bf182] > /usr/libexec/mysqld(_Z13handle_selectP3THDP3LEXP13select_resultm+0x184)[0= x5c5074] > /usr/libexec/mysqld[0x57df97] > /usr/libexec/mysqld(_Z21mysql_execute_commandP3THD+0x2438)[0x585808] > /usr/libexec/mysqld(_Z11mysql_parseP3THDPcjP12Parser_state+0x186)[0x589ef= 6] > /usr/libexec/mysqld(_Z16dispatch_command19enum_server_commandP3THDPcj+0x1= 5e5)[0x58b505] > /usr/libexec/mysqld(_Z24do_handle_one_connectionP3THD+0x117)[0x61fff7] > /usr/libexec/mysqld(handle_one_connection+0x50)[0x6200a0] > /lib64/libpthread.so.0(+0x6ccb)[0x7ffdaae8bccb] > /lib64/libc.so.6(clone+0x6d)[0x7ffda91bdc2d] > > Trying to get some variables. > Some pointers may be invalid and cause the dump to abort. > Query (0x7ffd20021720): is an invalid pointer > Connection ID (thread ID): 647 > Status: NOT_KILLED > > > -------- Original-Nachricht -------- > Betreff: 'myisam_use_mmap' unstable like hell > Datum: Thu, 24 Nov 2011 09:20:28 +0100 > Von: Reindl Harald > Organisation: the lounge interactive design > An: Mailing-List mysql > > introduced with 5.1 myisam_use_mmap leads in 5.5.18 > after some days to table crashes - will this be ever > useful on servers with thousands of tables? > ________________________ > > 111124 =A08:20:17 - 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 buil= t, > or misconfigured. This error can also be caused by malfunctioning hardwar= e. > We will try our best to scrape up some info that will hopefully help diag= nose > the problem, but since we have already crashed, something is definitely w= rong > and this may fail. > > key_buffer_size=3D268435456 > read_buffer_size=3D262144 > max_used_connections=3D33 > max_threads=3D200 > thread_count=3D6 > connection_count=3D6 > It is possible that mysqld could use up to > key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads =3D 4= 18015 K > bytes of memory > Hope that's ok; if not, decrease some variables in the equation. > > Thread pointer: 0x2a4bd50 > 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 =3D 0x7f507cf70d40 thread_stack 0x40000 > /usr/libexec/mysqld(my_print_stacktrace+0x33)[0x7ab8f3] > /usr/libexec/mysqld(handle_segfault+0x470)[0x50f190] > /lib64/libpthread.so.0(+0xeeb0)[0x7f51054caeb0] > /lib64/libc.so.6(+0x12ffa5)[0x7f5103843fa5] > /usr/libexec/mysqld(mi_mmap_pread+0x15a)[0x90880a] > /usr/libexec/mysqld(_mi_read_dynamic_record+0x1fe)[0x90ac5e] > /usr/libexec/mysqld(mi_rkey+0x378)[0x930f48] > /usr/libexec/mysqld(_ZN9ha_myisam14index_read_mapEPhPKhm16ha_rkey_functio= n+0x59)[0x8f1fe9] > /usr/libexec/mysqld[0x5b3f35] > /usr/libexec/mysqld(_Z10sub_selectP4JOINP13st_join_tableb+0x61)[0x5a4721] > /usr/libexec/mysqld[0x5b2c65] > /usr/libexec/mysqld(_ZN4JOIN4execEv+0xbe1)[0x5c39b1] > /usr/libexec/mysqld(_Z12mysql_selectP3THDPPP4ItemP10TABLE_LISTjR4ListIS1_= ES2_jP8st_orderSB_S2_SB_yP13select_resultP18st_select_lex_unitP13st_select_= lex+0x152)[0x5bf182] > /usr/libexec/mysqld(_Z13handle_selectP3THDP3LEXP13select_resultm+0x184)[0= x5c5074] > /usr/libexec/mysqld[0x57df97] > /usr/libexec/mysqld(_Z21mysql_execute_commandP3THD+0x2438)[0x585808] > /usr/libexec/mysqld(_Z11mysql_parseP3THDPcjP12Parser_state+0x186)[0x589ef= 6] > /usr/libexec/mysqld(_Z16dispatch_command19enum_server_commandP3THDPcj+0x1= 5e5)[0x58b505] > /usr/libexec/mysqld(_Z24do_handle_one_connectionP3THD+0x117)[0x61fff7] > /usr/libexec/mysqld(handle_one_connection+0x50)[0x6200a0] > /lib64/libpthread.so.0(+0x6ccb)[0x7f51054c2ccb] > /lib64/libc.so.6(clone+0x6d)[0x7f51037f4c2d] > > Trying to get some variables. > Some pointers may be invalid and cause the dump to abort. > Query (0x7f4ffc0058a0): is an invalid pointer > Connection ID (thread ID): 460043 > Status: NOT_KILLED > > The manual page at http://dev.mysql.com/doc/mysql/en/crashing.html contai= ns > information that should help you find out what is causing the crash. > > The "--memlock" argument, which was enabled, uses system calls that are > unreliable and unstable on some operating systems and operating-system > versions (notably, some versions of Linux). =A0This crash could be due to= use > of those buggy OS calls. =A0You should consider whether you really need t= he > "--memlock" parameter and/or consult the OS distributer about "mlockall" > bugs. > 111124 08:20:18 mysqld_safe Number of processes running now: 0 > 111124 08:20:18 mysqld_safe mysqld restarted > 111124 =A08:20:18 [Note] Plugin 'InnoDB' is disabled. > 111124 =A08:20:18 [Note] Plugin 'FEDERATED' is disabled. > 111124 =A08:20:18 [Note] Plugin 'BLACKHOLE' is disabled. > 111124 =A08:20:18 [Note] Plugin 'ARCHIVE' is disabled. > 111124 =A08:20:18 [Note] Plugin 'partition' is disabled. > 111124 =A08:20:19 [Note] Event Scheduler: Loaded 0 events > 111124 =A08:20:19 [Note] /usr/libexec/mysqld: ready for connections. > Version: '5.5.18-log' =A0socket: '/var/lib/mysql/mysql.sock' =A0port: 330= 6 =A0thelounge.net build > 111124 =A08:20:19 [ERROR] /usr/libexec/mysqld: Table './afi/cms1_global_c= ache' is marked as crashed and should be > repaired > 111124 =A08:20:19 [Warning] Checking table: =A0 './afi/cms1_global_cache' > 111124 =A08:20:19 [ERROR] Got an error from unknown thread, > /home/builduser/rpmbuild/BUILD/mysql-5.5.18/storage/myisam/ha_myisam.cc:8= 70 > > > > -- > > Mit besten Gr=FC=DFen, Reindl Harald > the lounge interactive design GmbH > A-1060 Vienna, Hofm=FChlgasse 17 > CTO / software-development / cms-solutions > p: +43 (1) 595 3999 33, m: +43 (676) 40 221 40 > icq: 154546673, http://www.thelounge.net/ > > http://www.thelounge.net/signature.asc.what.htm > --=20 Rob Wultsch wultsch@stripped