pbxt-discuss team mailing list archive

[Tim Soderstrom <tim@xxxxxxxxxxxxxxxxxxxxx>]

Hi Paul!

It may take some work to get a test case for the first error but I will try to get you both here soon.

Thanks!

Tim S.

On Apr 18, 2010, at 10:17 PM, Paul McCullagh wrote:

> Hi Tim,
> 
> On Apr 18, 2010, at 3:31 PM, Tim Soderstrom wrote:
> 
>> Hi PBXTers!
>> 
>> I am having trouble getting PBXT working properly on various versions of MySQL. When using MariaDB 5.1.42 and 5.1.44 with their built-in PBXT, it looks like PBXT gets stuck trying to replay logs? Here is an except of the error log:
>> 
>> Apr 18 22:28:04 mcp mysqld: 100418 22:28:04 [Error] DB-RECOVERY-mysql pthread_rwlock_t* xt_xlock_rwlock(thread_xt.cc:1644) errno (35): Resource deadlock avoided
>> Apr 18 22:28:04 mcp mysqld: 100418 22:28:04 [Error] DB-RECOVERY-mysql pthread_rwlock_t* xt_xlock_rwlock(thread_xt.cc:1644)
> 
> How difficult is it to repeat this error? Please report this as a bug, and add all the details you have. Thanks!
> 
>> For Percona 5.1.45, I end up getting a crash when trying to create a new table:
> 
> Please report this one too!
> 
>> 
>> ---
>> mysql> use test;
>> ERROR 1049 (42000): Unknown database 'test'
>> mysql> create database test;
>> Query OK, 1 row affected (0.00 sec)
>> 
>> mysql> use test;
>> Database changed
>> mysql> create table pbxt1 (num1 serial, string varchar(64)) ENGINE='PBXT';
>> ERROR 2013 (HY000): Lost connection to MySQL server during query
>> 
>> ..
>> 
>> mysql> Apr 18 22:30:52 mcp-new mysqld: 100418 22:30:52 - mysqld got signal 11 ;
>> Apr 18 22:30:52 mcp-new mysqld: This could be because you hit a bug. It is also possible that this binary
>> Apr 18 22:30:52 mcp-new mysqld: or one of the libraries it was linked against is corrupt, improperly built,
>> Apr 18 22:30:52 mcp-new mysqld: or misconfigured. This error can also be caused by malfunctioning hardware.
>> Apr 18 22:30:52 mcp-new mysqld: We will try our best to scrape up some info that will hopefully help diagnose
>> Apr 18 22:30:52 mcp-new mysqld: the problem, but since we have already crashed, something is definitely wrong
>> Apr 18 22:30:52 mcp-new mysqld: and this may fail.
>> Apr 18 22:30:52 mcp-new mysqld:
>> Apr 18 22:30:52 mcp-new mysqld: key_buffer_size=16777216
>> Apr 18 22:30:52 mcp-new mysqld: read_buffer_size=131072
>> Apr 18 22:30:52 mcp-new mysqld: max_used_connections=1
>> Apr 18 22:30:52 mcp-new mysqld: max_threads=151
>> Apr 18 22:30:52 mcp-new mysqld: threads_connected=1
>> Apr 18 22:30:52 mcp-new mysqld: It is possible that mysqld could use up to
>> Apr 18 22:30:52 mcp-new mysqld: key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 346495 K
>> Apr 18 22:30:52 mcp-new mysqld: bytes of memory
>> Apr 18 22:30:52 mcp-new mysqld: Hope that's ok; if not, decrease some variables in the equation.
>> Apr 18 22:30:52 mcp-new mysqld:
>> Apr 18 22:30:52 mcp-new mysqld: thd: 0x26c3850
>> Apr 18 22:30:52 mcp-new mysqld: Attempting backtrace. You can use the following information to find out
>> Apr 18 22:30:52 mcp-new mysqld: where mysqld died. If you see no messages after this, something went
>> Apr 18 22:30:52 mcp-new mysqld: terribly wrong...
>> Apr 18 22:30:52 mcp-new mysqld: stack_bottom = 0x7f27aeda30c8 thread_stack 0x20000
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(my_print_stacktrace+0x2e)[0x8dd90e]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(handle_segfault+0x40d)[0x5e104d]
>> Apr 18 22:30:52 mcp-new mysqld: /lib/libpthread.so.0[0x7f27c8f38190]
>> Apr 18 22:30:52 mcp-new mysqld: /lib/libc.so.6(strlen+0x11)[0x7f27c7f3c441]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/lib/mysql/plugin/libpbxt.so(_Z23xt_last_2_names_of_pathPKc+0x9)[0x7f27c4c46879]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/lib/mysql/plugin/libpbxt.so(_ZN18XTSystemTableShare13isSystemTableEPKc+0x1b)[0x7f27c4c59bab]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/lib/mysql/plugin/libpbxt.so[0x7f27c4c28a22]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(_Z15get_new_handlerP14st_table_shareP11st_mem_rootP10handlerton+0x68)[0x6d4228]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(_Z14open_table_defP3THDP14st_table_sharej+0x1359)[0x638709]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(_Z15ha_create_tableP3THDPKcS2_S2_P24st_ha_create_informationb+0x58)[0x6d7828]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(_Z16rea_create_tableP3THDPKcS2_S2_P24st_ha_create_informationR4ListI12Create_fieldEjP6st_keyP7handler+0x17b)[0x69791b]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(_Z26mysql_create_table_no_lockP3THDPKcS2_P24st_ha_create_informationP10Alter_infobj+0xc49)[0x6ec3a9]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(_Z18mysql_create_tableP3THDPKcS2_P24st_ha_create_informationP10Alter_infobj+0x8e)[0x6ec4fe]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(_Z21mysql_execute_commandP3THD+0x544a)[0x5f77da]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(_Z11mysql_parseP3THDPKcjPS2_+0x342)[0x5f7b32]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(_Z16dispatch_command19enum_server_commandP3THDPcj+0xd89)[0x5f88c9]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(_Z10do_commandP3THD+0xe8)[0x5f91a8]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(handle_one_connection+0x21d)[0x5eb8ed]
>> Apr 18 22:30:52 mcp-new mysqld: /lib/libpthread.so.0[0x7f27c8f2fa04]
>> Apr 18 22:30:52 mcp-new mysqld: /lib/libc.so.6(clone+0x6d)[0x7f27c7f9c80d]
>> Apr 18 22:30:52 mcp-new mysqld: Trying to get some variables.
>> Apr 18 22:30:52 mcp-new mysqld: Some pointers may be invalid and cause the dump to abort...
>> Apr 18 22:30:52 mcp-new mysqld: thd->query at 0x2bbbd90 = create table pbxt1 (num1 serial, string varchar(64)) ENGINE='PBXT'
>> Apr 18 22:30:52 mcp-new mysqld: thd->thread_id=35
>> Apr 18 22:30:52 mcp-new mysqld: thd->killed=NOT_KILLED
>> Apr 18 22:30:52 mcp-new mysqld: The manual page at http://dev.mysql.com/doc/mysql/en/crashing.html contains
>> Apr 18 22:30:52 mcp-new mysqld: information that should help you find out what is causing the crash.
>> Apr 18 22:30:52 mcp-new mysqld_safe: Number of processes running now: 0
>> Apr 18 22:30:52 mcp-new mysqld_safe: mysqld restarted
>> Apr 18 22:30:52 mcp-new mysqld: 100418 22:30:52 [Note] Plugin 'FEDERATED' is disabled.
>> Apr 18 22:30:52 mcp-new mysqld: InnoDB: The InnoDB memory heap is disabled
>> Apr 18 22:30:52 mcp-new mysqld: InnoDB: Mutexes and rw_locks use GCC atomic builtins
>> Apr 18 22:30:52 mcp-new mysqld: 100418 22:30:52  InnoDB: highest supported file format is Barracuda.
>> Apr 18 22:30:52 mcp-new mysqld: InnoDB: The log sequence number in ibdata files does not match
>> Apr 18 22:30:52 mcp-new mysqld: InnoDB: the log sequence number in the ib_logfiles!
>> Apr 18 22:30:52 mcp-new mysqld: 100418 22:30:52  InnoDB: Database was not shut down normally!
>> Apr 18 22:30:52 mcp-new mysqld: InnoDB: Starting crash recovery.
>> Apr 18 22:30:52 mcp-new mysqld: InnoDB: Reading tablespace information from the .ibd files...
>> Apr 18 22:30:52 mcp-new mysqld: InnoDB: Restoring possible half-written data pages from the doublewrite
>> Apr 18 22:30:52 mcp-new mysqld: InnoDB: buffer...
>> Apr 18 22:30:52 mcp-new mysqld: 100418 22:30:52 InnoDB Plugin 1.0.6-10 started; log sequence number 44278
>> Apr 18 22:30:52 mcp-new mysqld: 100418 22:30:52 [Note] PrimeBase XT (PBXT) Engine 1.0.10i RC loaded...
>> Apr 18 22:30:52 mcp-new mysqld: 100418 22:30:52 [Note] Paul McCullagh, PrimeBase Technologies GmbH, http://www.primebase.org
>> Apr 18 22:30:52 mcp-new mysqld: 100418 22:30:52 [Note] The server was not shutdown correctly, recovery required
>> Apr 18 22:30:52 mcp-new mysqld: 100418 22:30:52 [Note] Recovering after a crash using tc.log
>> Apr 18 22:30:52 mcp-new mysqld: 100418 22:30:52 - mysqld got signal 11 ;
>> Apr 18 22:30:52 mcp-new mysqld: This could be because you hit a bug. It is also possible that this binary
>> Apr 18 22:30:52 mcp-new mysqld: or one of the libraries it was linked against is corrupt, improperly built,
>> Apr 18 22:30:52 mcp-new mysqld: or misconfigured. This error can also be caused by malfunctioning hardware.
>> Apr 18 22:30:52 mcp-new mysqld: We will try our best to scrape up some info that will hopefully help diagnose
>> Apr 18 22:30:52 mcp-new mysqld: the problem, but since we have already crashed, something is definitely wrong
>> Apr 18 22:30:52 mcp-new mysqld: and this may fail.
>> Apr 18 22:30:52 mcp-new mysqld:
>> Apr 18 22:30:52 mcp-new mysqld: key_buffer_size=16777216
>> Apr 18 22:30:52 mcp-new mysqld: read_buffer_size=131072
>> Apr 18 22:30:52 mcp-new mysqld: max_used_connections=0
>> Apr 18 22:30:52 mcp-new mysqld: max_threads=151
>> Apr 18 22:30:52 mcp-new mysqld: threads_connected=0
>> Apr 18 22:30:52 mcp-new mysqld: It is possible that mysqld could use up to
>> Apr 18 22:30:52 mcp-new mysqld: key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 346495 K
>> Apr 18 22:30:52 mcp-new mysqld: bytes of memory
>> Apr 18 22:30:52 mcp-new mysqld: Hope that's ok; if not, decrease some variables in the equation.
>> Apr 18 22:30:52 mcp-new mysqld:
>> Apr 18 22:30:52 mcp-new mysqld: thd: 0x0
>> Apr 18 22:30:52 mcp-new mysqld: Attempting backtrace. You can use the following information to find out
>> Apr 18 22:30:52 mcp-new mysqld: where mysqld died. If you see no messages after this, something went
>> Apr 18 22:30:52 mcp-new mysqld: terribly wrong...
>> Apr 18 22:30:52 mcp-new mysqld: stack_bottom = (nil) thread_stack 0x20000
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(my_print_stacktrace+0x2e)[0x8dd90e]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(handle_segfault+0x40d)[0x5e104d]
>> Apr 18 22:30:52 mcp-new mysqld: /lib/libpthread.so.0[0x7fcdd8900190]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(_ZN11TC_LOG_MMAP7recoverEv+0x149)[0x67bd99]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(_ZN11TC_LOG_MMAP4openEPKc+0x3f8)[0x67e678]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld[0x5e30c8]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld(main+0x1de)[0x5e61ce]
>> Apr 18 22:30:52 mcp-new mysqld: /lib/libc.so.6(__libc_start_main+0xfd)[0x7fcdd78a3abd]
>> Apr 18 22:30:52 mcp-new mysqld: /usr/sbin/mysqld[0x516aa9]
>> Apr 18 22:30:52 mcp-new mysqld: The manual page at http://dev.mysql.com/doc/mysql/en/crashing.html contains
>> Apr 18 22:30:52 mcp-new mysqld: information that should help you find out what is causing the crash.
>> Apr 18 22:30:52 mcp-new mysqld_safe: mysqld from pid file /var/run/mysqld/mysqld.pid ended
>> ---
>> 
>> I have not logged a bug yet since I wasn't sure if there was enough information to log one but thought I would at least pass this along.
> 
> It sounds like this error is easy to repeat, so please to report it.
> 
> Thanks,
> 
> Paul
> 
> --
> Paul McCullagh
> PrimeBase Technologies
> www.primebase.org
> www.blobstreaming.org
> pbxt.blogspot.com
> 
> 
>