Re: mysqld got signal 11; This could be because you hit a bug ....
Hello, I have followed your suggestion. I have installed mysql-debug-4.0.10-gamma-sun-solaris2.9-sparc-64bit on the Solaris 8 (Machine with 8 cpus, 32GB of Memory and 64 of swap) server. I have started the mysqld program as root I have specified the core-file in the /etc/my.cnf file. Other info that could be interesting in the /etc/my.cnf are _ set-variable = max_connections=16 set-variable = max_tmp_tables=512 set-variable= table_cache=512 set-variable = key_buffer=16384M set-variable= sort_buffer=1K set-variable= record_buffer=1K set-variable= thread_cache_size=8 set-variable= thread_cache=512 set-variable= thread_concurrency=12 _ I have started 7 concurrent mysql threads (7 OPTIMIZE TABLE from mysql command line) and after 9 hours I got a segmentation fault with Error ERROR 2013: Lost connection to MySQL server during query ERROR 2013: Lost connection to MySQL server during query ERROR 2013: Lost connection to MySQL server during query ERROR 2013: Lost connection to MySQL server during query ERROR 2013: Lost connection to MySQL server during query ERROR 2013: Lost connection to MySQL server during query ERROR 2013: Lost connection to MySQL server during query Segmentation Fault 030223 16:11:05 mysqld restarted I have looked at the mysqld error log file and this is the message 030223 07:03:11 mysqld started /usr/local/mysql-debug-4.0.10-gamma-sun-solaris2.9-sparc-64bit/bin/mysqld: ready for connections. Version: '4.0.10-gamma-debug-debug-log' socket: '/tmp/mysql.sock' port: 3306 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. 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. key_buffer_size=17179869184 read_buffer_size=8192 sort_buffer_size=32768 max_used_connections=7 max_connections=16 threads_connected=8 It is possible that mysqld could use up to key_buffer_size + (read_buffer_size + sort_buffer_size)*max_connections = 16777856 K bytes of memory Hope that's ok; if not, decrease some variables in the equation. Writing a core file 030223 16:11:05 mysqld restarted __ I have looked for the core file, but I couldn't find it, even though in the error log it says Writing a core file. Does this give you any clue about what is causing the error ? Could you help please about the core file Do I need to specify the option differently ? Thanks in advance for your help, Mariella At 04:34 PM 1/29/03 +0200, Sinisa Milivojevic wrote: On Tue, 28 Jan 2003 17:12:14 -0700 Mariella Di Giacomo [EMAIL PROTECTED] wrote: Hello, I am running mysql server (mysql-standard-4.0.5-beta-64bit) 64 bit on Sun Solaris 2.8 and I have got the binaries form the mysql web site. The machine I am using has 32GB of RAM + 64GB of swap. Thanks in advance for your help, Mariella Hi! The reason why MySQL is crashing is not due to the fact that your memory is exhausted. It is possible that you have shell limits that are far below the available memory, but in your case even that is not a problem. What you should do is get a 4.0.10 debug binary for Solaris 2.8 from our site and send us a core file that is created on the crash. Also add --core-file to startup options. -- Regards, -- For technical support contracts, go to https://order.mysql.com/?ref=msmi __ ___ ___ __ / |/ /_ __/ __/ __ \/ /Mr. Sinisa Milivojevic [EMAIL PROTECTED] / /|_/ / // /\ \/ /_/ / /__ MySQL AB, FullTime Developer /_/ /_/\_, /___/\___\_\___/ Larnaca, Cyprus ___/ www.mysql.com - Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail [EMAIL PROTECTED] To unsubscribe, e-mail [EMAIL PROTECTED] Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
Re: mysqld got signal 11; This could be because you hit a bug ....
Hi, I will install the 2.8 debug version. The reason why I set setting sort and record buffers to 1K is because with a few tests I have noticed that I was getting a better response time for queries. The limit for the core file was unlimited. I will look into customer support. Mariella At 06:17 PM 2/24/03 +0200, you wrote: Mariella Di Giacomo writes: Hello, I have followed your suggestion. I have installed mysql-debug-4.0.10-gamma-sun-solaris2.9-sparc-64bit on the Solaris 8 (Machine with 8 cpus, 32GB of Memory and 64 of swap) server. I have started the mysqld program as root I have specified the core-file in the /etc/my.cnf file. Other info that could be interesting in the /etc/my.cnf are HI! Causes of the crashes could be many, among the others: * using 2.9 binary on Solaris 2.8 !! Please use 2.8 binary, we have it * setting sort and record buffers to 1K !! Please set them at 1M There are many other things that could be improved, but that can be done when and if you become a registered customer. I have looked for the core file, but I couldn't find it, even though in the error log it says Writing a core file. Make sure all ulimit values are OK, including the one for core file. Does this give you any clue about what is causing the error ? Could you help please about the core file Do I need to specify the option differently ? Thanks in advance for your help, Mariella Regards, -- For technical support contracts, go to https://order.mysql.com/?ref=msmi __ ___ ___ __ / |/ /_ __/ __/ __ \/ /Mr. Sinisa Milivojevic [EMAIL PROTECTED] / /|_/ / // /\ \/ /_/ / /__ MySQL AB, Fulltime Developer /_/ /_/\_, /___/\___\_\___/ Larnaca, Cyprus ___/ www.mysql.com Join MySQL Users Conference and Expo: http://www.mysql.com/events/uc2003/ - Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail [EMAIL PROTECTED] To unsubscribe, e-mail [EMAIL PROTECTED] Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
Re: mysqld got signal 11; This could be because you hit a bug ....
Hi, I would prefer the debug version which the latest I believe is the 4.0.10. The same error I get it when I run ALTER TABLES or I try to insert data into tables. I will try and let you know. Thanks, Mariella At 06:31 PM 2/24/03 +0200, you wrote: Mariella Di Giacomo writes: Hi, I will install the 2.8 debug version. The reason why I set setting sort and record buffers to 1K is because with a few tests I have noticed that I was getting a better response time for queries. The limit for the core file was unlimited. I will look into customer support. Mariella Please also make sure that mysqld is run under mysql user and that all UNIX permissions are fine. BTW 4.0.11 is out with some fixes for OPTIMIZE table. Regards, -- For technical support contracts, go to https://order.mysql.com/?ref=msmi __ ___ ___ __ / |/ /_ __/ __/ __ \/ /Mr. Sinisa Milivojevic [EMAIL PROTECTED] / /|_/ / // /\ \/ /_/ / /__ MySQL AB, FullTime Developer /_/ /_/\_, /___/\___\_\___/ Larnaca, Cyprus ___/ www.mysql.com - Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail [EMAIL PROTECTED] To unsubscribe, e-mail [EMAIL PROTECTED] Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
Does MySQL store TimeZone and DayLight Saving Time changes ?
Hello, We are running MySQL server Ver 3.23.8 on a Linux system. We have a table that contains a field whose type is MODIFIED | timestamp(14) We would like to find out if there an easy way in MySQL to convert the date that is on the MODIFIED field (timestamp) to GMT time without an external program. Does MySQL store TimeZone and DayLight Saving Time changes ? Or does it have a way to find it once you have a field whose type is timestamp ? Thanks a lot in advance for your help, Patrick, Mariella - Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail [EMAIL PROTECTED] To unsubscribe, e-mail [EMAIL PROTECTED] Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
mysqld got signal 11; This could be because you hit a bug ....
Hello, I am running mysql server (mysql-standard-4.0.5-beta-64bit) 64 bit on Sun Solaris 2.8 and I have got the binaries form the mysql web site. The machine I am using has 32GB of RAM + 64GB of swap. At the certain point I got : __ /usr/local/mysql-standard-4.0.5-beta-64bit/bin/mysqld: ready for connections 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. 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. key_buffer_size=11274285056 read_buffer_size=2093056 sort_buffer_size=2097144 max_used_connections=0 max_connections=16 threads_connected=1 It is possible that mysqld could use up to key_buffer_size + (read_buffer_size + sort_buffer_size)*max_connections = 11075515 K bytes of memory Hope that's ok; if not, decrease some variables in the equation. __ I have tried also 4.0.8 and I have the same problem. I do not think it is an hardware problem, because I have tried the same task on a different Sun machine (with Solaris 2.8 and the same executable) and I get the same problem after a while. I have tested the memory allocation with mysqld through the creation of big HEAP table and mysqld could allocate all the memory (32GB+64GB) for the heap table. I have got an error after that because mysqld needed more memory for finishing filling the heap table and it could not allocate more (xmalloc failed) of it. So if mysqld can use all the Solaris Virtual memory of the system, why do I get that problem, provided that I have more memory (32GB + 64GB) of what it is complaining ? key_buffer_size + (read_buffer_size + sort_buffer_size)*max_connections = 11075515 K Could someone help, please ? Thanks in advance for your help, Mariella - Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail [EMAIL PROTECTED] To unsubscribe, e-mail [EMAIL PROTECTED] Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
Re: mysqld got signal 11; This could be because you hit a bug ....
Hi Mark, Thanks. The memory that we are using is ECC. The reason I have only 16 connections is because that DB is used only for updates. In reality for that DB we use 3 or 4 connections maximum. We have a mirror copy used for customers which has a lot of connections. Mariella At 10:22 AM 1/29/03 +0100, Mark wrote: - Original Message - From: Mariella Di Giacomo [EMAIL PROTECTED] To: [EMAIL PROTECTED]; [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Sent: Wednesday, January 29, 2003 1:12 AM Subject: mysqld got signal 11; This could be because you hit a bug Hello, I am running mysql server (mysql-standard-4.0.5-beta-64bit) 64 bit on Sun Solaris 2.8 and I have got the binaries form the mysql web site. The machine I am using has 32GB of RAM + 64GB of swap. Hi Mariella, My, that is a LOT of memory! :) What could you possibly have running that allows only 16 connections, yet uses up to the max of 96 GigaByte? :) Yikes! I take it this is high-quality ECC memory? If not, with that much memory, a seg fault could easily occur. I believe FreeBSD already advises to uses ECC memory when you have as little as 4 G. I'm sure you got that covered, though; just checking. :) - Mark - Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail [EMAIL PROTECTED] To unsubscribe, e-mail [EMAIL PROTECTED] Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
Slow response with MRG tables.
Hello, I would like to ask three questions related to MySQL and MERGE tables. I have a program that reads from mysql server (4.0.5 for Solaris 2.8 on a V880 with 8 processors and 32 GB of memory), using Perl DBI MySQl driver, two tables FULLM (merge of 3 tables) and RECKM (merge of 15 tables), that are MERGED TABLES. There is an update program that reads (it has been running for weeks) both tables doing a LEFT JOIN in the following matter, where the value of vkeys changes for each query: select substring(collection,1,3) as coll, count(*) from FULLM left join RECKM on intkeys=starkey where vkeys= /cors/abcd00/2301-4598/1970/0/79_RODBARD-CARARD group by coll; I have the slow log option turned on and in general these queries take a long time (in most cases over a minute). What I have noticed is that most of the time is spent in querying the RECKM merged table (even in the case where I do not have a join), no matter if I am querying through Perl or directly from mysql command line. My questions are : 1) when a query is made to a MRG table, how is that query implemented by MySQL ? 2) Is there anything I could optimize in the query itself ? I cannot have ENUM Type fields as procedure analyse() suggests. 3) I have included the table types and description, the mysqld conf file. Could someone help please ? Thanks a lot in advance for your help, Mariella _ SELECT substring(collection, 1, 3) as coll, count(*) from FULLM LEFT JOIN RECKM on intkeys=starkey where vkeys= /cors/abcd00/0003-0147/106/949/283_NEI-GDBP group by coll PROCEDURE ANALYSE() \G *** 1. row *** Field_name: coll Min_value: Art Max_value: Art Min_length: 3 Max_length: 3 Empties_or_zeros: 0 Nulls: 0 Avg_value_or_avg_length: 3. Std: NULL Optimal_fieldtype: ENUM('Art') NOT NULL *** 2. row *** Field_name: count(*) Min_value: 4 Max_value: 3628 Min_length: 1 Max_length: 4 Empties_or_zeros: 0 Nulls: 0 Avg_value_or_avg_length: 1250. Std: 1681.9196 Optimal_fieldtype: ENUM('4','119','3628') NOT NULL 2 rows in set (0.78 sec) _ mysql describe FULLM; +-+--+--+-+-+---+ | Field | Type | Null | Key | Default | Extra | +-+--+--+-+-+---+ | vkeys | varchar(100) | | PRI | | | | intkeys | varchar(11) | YES | | NULL | | +-+--+--+-+-+---+ 2 rows in set (0.01 sec) mysql describe RECKM; ++--+--+-+-+---+ | Field | Type | Null | Key | Default | Extra | ++--+--+-+-+---+ | vdkkey | varchar(100) | YES | MUL | NULL | | | starkey | varchar(11) | YES | MUL | NULL | | | collection | varchar(15) | YES | | NULL | | ++--+--+-+-+---+ 3 rows in set (0.00 sec) show index FROM RECKM \G *** 1. row *** Table: RECKM Non_unique: 1 Key_name: index_vdkkey Seq_in_index: 1 Column_name: vdkkey Collation: A Cardinality: NULL Sub_part: NULL Packed: NULL Null: YES Index_type: Comment: *** 2. row *** Table: RECKM Non_unique: 1 Key_name: index_starkey Seq_in_index: 1 Column_name: starkey Collation: A Cardinality: NULL Sub_part: NULL Packed: NULL Null: YES Index_type: Comment: 2 rows in set (0.00 sec) show index FROM FULLM \G *** 1. row *** Table: FULLM Non_unique: 0 Key_name: PRIMARY Seq_in_index: 1 Column_name: vkeys Collation: A Cardinality: NULL Sub_part: NULL Packed: NULL Null: Index_type: Comment: 1 row in set (0.00 sec) show table STATUS like FULLM \G *** 1. row *** Name: FULLM Type: MRG_MyISAM Row_format: Dynamic Rows: 17610691 Avg_row_length: 112 Data_length: 1348904896 Max_data_length: NULL Index_length: 0 Data_free: 0 Auto_increment: NULL Create_time: NULL Update_time: NULL Check_time: NULL Create_options: Comment: show table STATUS like RECKEYS \G *** 1. row *** Name: RECKEYS Type: MRG_MyISAM Row_format: Dynamic Rows: 491596360 Avg_row_length: 127 Data_length: 40240090580 Max_data_length: NULL Index_length: 0 Data_free: 0 Auto_increment: NULL Create_time: NULL Update_time: NULL Check_time: NULL Create_options: Comment: This is the mysqld my.cnf file # set-variable = max_connections=16384 set-variable =
Question about MySQL and lwp_mutex_lock and lwp_mutex_wakeup
Hello, I am running MySQL 4.0.5b 64 bits on a Solaris 2.8. The server used has 8 CPUs. MySQL DB has access (and menage) in that system 2DBs. The first DB is only is accessed only through SELECTS and the second one has been used for UPDATES and INSERTS. Initially I was running a process that was updating the tables. In the program there is no contention among tables and there is only one process that executes UPDATES AND INSERTS. Apparently as soon as people start accessing the other DB (quite heavily) the time spent by MySQL in lwp_mutex_wakeup lwp_mutex_lock increases a lot . This is a screen shot taken with truss of the process in a certain timeframe. __ signals SIGALRM1 total: 1 syscall seconds calls errors read 69.784597594 write 26.411851 close.46 105 time 87.915096 lseek 6.44 276 alarm.00 1 fstat 17.101503 access .72 105 fcntl 22.591503 poll1.89 105 lwp_sema_wait 4.88 562 lwp_sema_post 7.95 562 lwp_mutex_wakeup 30.893247 lwp_mutex_lock 31.172529 lwp_cond_wait.00 1 1 pread 2.73 137 pwrite 45.362361 signotifywait.00 1 accept 2.43 105 shutdown .99 105 getpeername .73 105 getsockname 2.18 105 getsockopt 22.551503 setsockopt 26.881713 lwp_sigtimedwait .07 1 --- -- sys totals: 412.11 28179595 usr time:.48 elapsed: 96.46 __ As soon as all the shutdown all the users from the DB that is used for only SELECT, now the same process that runs MySQL updates on the other DB has hardly ever lwp_mutex_wakeup lwp_mutex_lock calls. Could someone, please, help me trying to understand why that is the behavior ? Thanks a lot, Mariella - Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail [EMAIL PROTECTED] To unsubscribe, e-mail [EMAIL PROTECTED] Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
enable-assembler error and MySQL 64 bits compilation error
X-Sender: [EMAIL PROTECTED] X-Mailer: QUALCOMM Windows Eudora Version 5.0 Date: Fri, 06 Dec 2002 09:20:55 -0700 To: [EMAIL PROTECTED] From: Mariella Di Giacomo [EMAIL PROTECTED] Subject: --enable-assembler error and MySQL 64 bits compilation error on Solaris 2.8 Hello, I have been running MySQL 4.0.5 64 bits on a Sun sun4u sparc SUNW,Sun-Fire-880, Solaris 2.8. I have been using the binary version that MySQL provides. Yesterday I have downloaded the Solaris tar ball to compile the MySQL source code and change the myisam block size. I have the Solaris compiler cc: Sun WorkShop (Fort) 6 update 2 C 5.3 2001/05/15 and I wanted to use that because it is suppose to create a faster executable than gcc. I have followed the instructions that are provided for the Solaris compilation and cc Forte which are CC=cc CFLAGS=-Xa -fast -native -xstrconst -mt -xarch=v9 \ CXX=CC CXXFLAGS=-noex -mt -xarch=v9 \ ./configure --prefix=/usr/local/mysql --enable-assembler At the certain point I get the following error: Making all in strings /usr/ccs/bin/as -O -DDBUG_OFF -Xa -fast -native -xstrconst -mt -xarch=v9 -D_FILE_OFFSET_BITS=64 -DHAVE_CURSES_H -I/root/mysql/mysql-4.0.5a-beta/include -DHAVE_RWLOCK_T -o bmove_upp-sparc.o bmove_upp-sparc.s /usr/ccs/bin/as: error: unknown option 'O' /usr/ccs/bin/as: error: unknown option 'X' /usr/ccs/bin/as: error: unknown option 'a' /usr/ccs/bin/as: error: unknown option 'f' /usr/ccs/bin/as: error: unknown option 'ast' /usr/ccs/bin/as: error: unknown option 'ative' /usr/ccs/bin/as: error: unknown option 'xstrconst' /usr/ccs/bin/as: error: unknown option 't' usage: /usr/ccs/bin/as [-V] [-Q{y,n}] [-f[O][if]#] [-q] [-s] [-S] [-K {pic,PIC}] [-o objfile] [-L] [-T] [-P [[-Ipath] [-Dname] [-Dname=def] [-Uname]]...] [-m [-Ym,path]] [-n] [-xF] [-xarch=v7] [-xarch=v8] [-xarch=v8a] [-xarch=v8plus] [-xarch=v8plusa] [-xarch=v8plusb] [-xarch=v9] [-xarch=v9a] [-xarch=v9b] [-xcode={pic13,pic32}] file.s... *** Error code 1 make: Fatal error: Command failed for target `bmove_upp-sparc.o' Current working directory /root/mysql/mysql-4.0.5a-beta/strings *** Error code 1 make: Fatal error: Command failed for target `all-recursive' Current working directory /root/mysql/mysql-4.0.5a-beta *** Error code 1 make: Fatal error: Command failed for target `all' /usr/ccs/bin/as does not have those options and I could not find a patch for as. I can get things compiled only If I do not specify --enable-assembler CC=cc CFLAGS=-Xa -fast -native -xstrconst -mt -xarch=v9 \ CXX=CC CXXFLAGS=-noex -mt -xarch=v9 \ ./configure --prefix=/usr/local/mysql So my questions: Do you know why ? Is there an error in the configure script ? or I cannot compile with --enable-assembler because in the system Use assembler versions of some string functions is not available ? If that is the case, is there any work around that I could use to compile with --enable-assembler ? If I cannot use --enable-assembler, which is the percentage of Performance I may lose ? Thanks a lot for your help, Mariella - Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail [EMAIL PROTECTED] To unsubscribe, e-mail [EMAIL PROTECTED] Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
Question about MySQL replication and OPTIMIZE TABLE
Hello, I am running MySQL 4.0.5 64 bit on a Sun Solaris and I am using MyISAM Indices. I would like to set replication (one master and one slave). My question is: When I OPTIMIZE TABLE table1, ... on the master will the slave be updated or if I want to optimize the same tables on the slave I need to explicitly run OPTIMIZE TABLE on the slave side ? Thanks a lot for your help, Mariella - Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail [EMAIL PROTECTED] To unsubscribe, e-mail [EMAIL PROTECTED] Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
Question about myisam_block_size
Hello, I am running MySQL 4.0.5 64 bit on a Sun Solaris and I am using MyISAM Indices. I have noticed that the myisam_block_size is set to 1K. What does it happen if I am going to change to 8K ? The reason for that is because the Solaris pagesize is set to 8K, so I was thinking that maybe I could get improvements in performance. Is that correct ? Thanks a lot in advance for your help, Mariella - Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail [EMAIL PROTECTED] To unsubscribe, e-mail [EMAIL PROTECTED] Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php