- discuss - lists.mariadb.org

[Maria-discuss] New Question: MariaDB on SGE with environmental modules
by AskMonty KB 28 Sep '15

28 Sep '15

Hello, A new question has been asked in "Getting, Installing, and Upgrading MariaDB" by netzach. Please answer it at http://mariadb.com/kb/en/mariadb-on-sge-with-environmental-modules/ as the person asking the question may not be subscribed to the mailing list. -------------------------------- Hi The title says it all. How do I setup MariaDB on a computer cluster using SGE and environmental modules with one shared place for the databases? I want the databases to be in /databases which is exported through a local network to the computational nodes. All computers run centos 7. Thank you Jonas Söderberg -------------------------------- To view or answer this question please visit: http://mariadb.com/kb/en/mariadb-on-sge-with-environmental-modules/

1 0

[Maria-discuss] MariaDB Crashes due to Stored Procedure
by Joseph Onyilo 27 Sep '15

27 Sep '15

I created two Store procedures to time table scheduling some months back and test it at the time which was fine. The first procedure has a cursor to navigate a list of records and call the second procedure. The Second Procedure has two cursors for different scenarios. Now trying to a final test before moving to production and it crashes my MariaDB Server mid way. Even when I called the Second Procedure individually with static parameters, after three to five calls MariaDB still crashes. The Operating system is Centos 6.6, MariaDB 10.0.21 and error from log displays is as follows |This could be because you hit a bug.It isalso possible that this binary orone ofthe libraries it was linked against iscorrupt,improperly built,ormisconfigured.This error can also be caused bymalfunctioning hardware.Toreport this bug,see http://kb.askmonty.org/en/reporting-bugs We will try our best toscrape up someinfo that will hopefully help diagnose the problem,but since we have already crashed,something isdefinitely wrong andthis may fail.Server version:10.0.21-MariaDB key_buffer_size=134217728read_buffer_size=131072max_used_connections=3max_threads=153thread_count=2It ispossible that mysqld could useup tokey_buffer_size +(read_buffer_size +sort_buffer_size)*max_threads =467003K bytes ofmemory Hope that's ok;ifnot,decrease somevariables inthe equation.Thread pointer:0x0x7fdac266f008 Attempting backtrace.You can usethe followinginformation tofind out wheremysqld died.Ifyou see no messages after this,something went terribly wrong...stack_bottom =0x7fdad74cfd30thread_stack 0x48000/usr/sbin/mysqld(my_print_stacktrace+0x2b)[0xb7f84b]/usr/sbin/mysqld(handle_fatal_signal+0x398)[0x72b758]/lib64/libpthread.so.0(+0xf710)[0x7fdad710b710]/usr/sbin/mysqld(_ZN13st_select_lex17mark_as_dependentEP3THDPS_P4Item+0xa0)[0x5c8410]/usr/sbin/mysqld[0x73c544]/usr/sbin/mysqld(_ZN10Item_field15fix_outer_fieldEP3THDPP5FieldPP4Item+0x62f)[0x74c78f]/usr/sbin/mysqld(_ZN10Item_field10fix_fieldsEP3THDPP4Item+0x432)[0x74ce72]/usr/sbin/mysqld(_Z12setup_fieldsP3THDPP4ItemR4ListIS1_E17enum_mark_columnsPS5_b+0x163)[0x594a03]/usr/sbin/mysqld[0x6168f3]/usr/sbin/mysqld(_ZN18st_select_lex_unit7prepareEP3THDP13select_resultm+0x39c)[0x6648ac]/usr/sbin/mysqld(_Z21mysql_derived_prepareP3THDP3LEXP10TABLE_LIST+0x206)[0x5b4bc6]/usr/sbin/mysqld(_Z27mysql_handle_single_derivedP3LEXP10TABLE_LISTj+0x193)[0x5b5bd3]/usr/sbin/mysqld(_ZN13st_select_lex14handle_derivedEP3LEXj+0x47)[0x5ca857]/usr/sbin/mysqld(_ZN10TABLE_LIST14handle_derivedEP3LEXj+0x4f)[0x679f2f]/usr/sbin/mysqld(_Z28mysql_handle_list_of_derivedP3LEXP10TABLE_LISTj+0x3f)[0x5b5d4f]/usr/sbin/mysqld(_Z20mysql_prepare_insertP3THDP10TABLE_LISTP5TABLER4ListI4ItemEPS7_S8_S8_15enum_duplicatesPPS6_bbb+0xb8)[0x5c09a8]/usr/sbin/mysqld(_Z27mysql_insert_select_prepareP3THD+0x72)[0x5c0e62]/usr/sbin/mysqld(_Z21mysql_execute_commandP3THD+0x556f)[0x5d94bf]/usr/sbin/mysqld(_ZN13sp_instr_stmt9exec_coreEP3THDPj+0x1e)[0x81d26e]/usr/sbin/mysqld(_ZN13sp_lex_keeper23reset_lex_and_exec_coreEP3THDPjbP8sp_instr+0x8b)[0x82285b]/usr/sbin/mysqld(_ZN13sp_instr_stmt7executeEP3THDPj+0x155)[0x822d45]/usr/sbin/mysqld(_ZN7sp_head7executeEP3THDb+0x6c9)[0x820529]/usr/sbin/mysqld(_ZN7sp_head17execute_procedureEP3THDP4ListI4ItemE+0x5f6)[0x821796]/usr/sbin/mysqld(_Z21mysql_execute_commandP3THD+0x31f9)[0x5d7149]/usr/sbin/mysqld(_ZN13sp_instr_stmt9exec_coreEP3THDPj+0x1e)[0x81d26e]/usr/sbin/mysqld(_ZN13sp_lex_keeper23reset_lex_and_exec_coreEP3THDPjbP8sp_instr+0x8b)[0x82285b]/usr/sbin/mysqld(_ZN13sp_instr_stmt7executeEP3THDPj+0x155)[0x822d45]/usr/sbin/mysqld(_ZN7sp_head7executeEP3THDb+0x6c9)[0x820529]/usr/sbin/mysqld(_ZN7sp_head17execute_procedureEP3THDP4ListI4ItemE+0x5f6)[0x821796]/usr/sbin/mysqld(_Z21mysql_execute_commandP3THD+0x31f9)[0x5d7149]/usr/sbin/mysqld[0x5da9d7]/usr/sbin/mysqld(_Z16dispatch_command19enum_server_commandP3THDPcj+0x1d0d)[0x5dce4d]/usr/sbin/mysqld(_Z24do_handle_one_connectionP3THD+0x453)[0x69a2d3]/usr/sbin/mysqld(handle_one_connection+0x42)[0x69a3a2]/lib64/libpthread.so.0(+0x79d1)[0x7fdad71039d1]/lib64/libc.so.6(clone+0x6d)[0x7fdad581d8fd]Trying toget somevariables.Somepointers may be invalid andcause the dumptoabort.Query (0x7fdaa8818548):isan invalid pointer Connection ID (thread ID):3Status: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,engine_condition_pushdown=off,index_condition_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,partial_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_incremental=on,join_cache_hashed=on,join_cache_bka=on,optimize_join_buffer_size=off,table_elimination=on,extended_keys=on,exists_to_in=onThe manual page at http://dev.mysql.com/doc/mysql/en/crashing.html containsinformation that should help you find out what iscausing the crash.| I also did a test on MariaDB 10.0.17 on windows still no luck.

1 1

[Maria-discuss] select using date_sub produces unwanted milliseconds in results
by Dan Ragle 26 Sep '15

26 Sep '15

We're converting from MySQL 5.1 to MariaDB 5.5.44 and I'm hoping someone can help clarify for me what's going on in the following scenario. In brief, we're getting milliseconds displayed in a particular date_sub calculation (and possibly other selects with date arithmetic, trying to figure out how to track them down is part of the goal of this post) where we aren't expecting them. MariaDB [testdb]> show variables like '%version%'; +-------------------------+---------------------+ | Variable_name | Value | +-------------------------+---------------------+ | innodb_version | 5.5.43-MariaDB-37.2 | | protocol_version | 10 | | slave_type_conversions | | | version | 5.5.44-MariaDB-log | | version_comment | MariaDB Server | | version_compile_machine | x86_64 | | version_compile_os | Linux | +-------------------------+---------------------+ 7 rows in set (0.00 sec) MariaDB [testdb]> show create table test \G *************************** 1. row *************************** Table: test Create Table: CREATE TABLE `test` ( `datetime` datetime NOT NULL ) ENGINE=MyISAM DEFAULT CHARSET=latin1 1 row in set (0.00 sec) MariaDB [testdb]> select datetime from test; Field 1: `datetime` Catalog: `def` Database: `velocitynorep` Table: `test` Org_table: `test` Type: DATETIME Collation: binary (63) Length: 19 Max_length: 19 Decimals: 0 Flags: NOT_NULL BINARY NO_DEFAULT_VALUE +---------------------+ | datetime | +---------------------+ | 2015-09-18 09:23:21 | +---------------------+ 1 row in set (0.00 sec) MariaDB [testdb]> select date_sub(datetime,interval 1 second) from test; Field 1: `date_sub(datetime,interval 1 second)` Catalog: `def` Database: `` Table: `` Org_table: `` Type: DATETIME Collation: binary (63) Length: 19 Max_length: 19 Decimals: 0 Flags: BINARY +--------------------------------------+ | date_sub(datetime,interval 1 second) | +--------------------------------------+ | 2015-09-18 09:23:20 | +--------------------------------------+ 1 row in set (0.00 sec) MariaDB [testdb]> select substr(datetime,1,13) from test; Field 1: `substr(datetime,1,13)` Catalog: `def` Database: `` Table: `` Org_table: `` Type: VAR_STRING Collation: utf8_general_ci (33) Length: 39 Max_length: 13 Decimals: 31 Flags: NOT_NULL +-----------------------+ | substr(datetime,1,13) | +-----------------------+ | 2015-09-18 09 | +-----------------------+ 1 row in set (0.00 sec) MariaDB [testdb]> select concat(substr(datetime,1,13),':00:00') -> from test; Field 1: `concat(substr(datetime,1,13),':00:00')` Catalog: `def` Database: `` Table: `` Org_table: `` Type: VAR_STRING Collation: utf8_general_ci (33) Length: 57 Max_length: 19 Decimals: 31 Flags: NOT_NULL +----------------------------------------+ | concat(substr(datetime,1,13),':00:00') | +----------------------------------------+ | 2015-09-18 09:00:00 | +----------------------------------------+ 1 row in set (0.00 sec) MariaDB [testdb]> select concat('2015-09-18 09',':00:00'); Field 1: `concat('2015-09-18 09',':00:00')` Catalog: `def` Database: `` Table: `` Org_table: `` Type: VAR_STRING Collation: utf8_general_ci (33) Length: 57 Max_length: 19 Decimals: 31 Flags: NOT_NULL +----------------------------------+ | concat('2015-09-18 09',':00:00') | +----------------------------------+ | 2015-09-18 09:00:00 | +----------------------------------+ 1 row in set (0.00 sec) MariaDB [testdb]> select date_sub(concat('2015-09-18 09',':00:00'), -> interval 1 second); Field 1: `date_sub(concat('2015-09-18 09',':00:00'), interval 1 second)` Catalog: `def` Database: `` Table: `` Org_table: `` Type: STRING Collation: binary (63) Length: 19 Max_length: 19 Decimals: 0 Flags: BINARY +---------------------------------------------------------------+ | date_sub(concat('2015-09-18 09',':00:00'), interval 1 second) | +---------------------------------------------------------------+ | 2015-09-18 08:59:59 | +---------------------------------------------------------------+ 1 row in set (0.00 sec) MariaDB [testdb]> select -> date_sub(concat(substr(datetime,1,13),':00:00'), -> interval 1 second) from test; Field 1: `date_sub(concat(substr(datetime,1,13),':00:00'), interval 1 second)` Catalog: `def` Database: `` Table: `` Org_table: `` Type: STRING Collation: binary (63) Length: 26 Max_length: 26 Decimals: 6 Flags: BINARY +---------------------------------------------------------------------+ | date_sub(concat(substr(datetime,1,13),':00:00'), interval 1 second) | +---------------------------------------------------------------------+ | 2015-09-18 08:59:59.000000 | +---------------------------------------------------------------------+ 1 row in set (0.00 sec) The last one is the one that has me stumped. 1.) Why does it include milliseconds in the output? According to this page: https://mariadb.com/kb/en/mariadb/microseconds-in-mariadb/ where no precision is specified the result should not display milliseconds, which is what happens in all the other examples above except this one. 2.) The statement on a 5.1 MySQL server (5.1.73-log, also 64 bit Linux) produces this: mysql> select date_sub(concat(substr(datetime,1,13),':00:00'), -> interval 1 second) from test; Field 1: `date_sub(concat(substr(datetime,1,13),':00:00'), interval 1 second)` Catalog: `def` Database: `` Table: `` Org_table: `` Type: STRING Collation: binary (63) Length: 29 Max_length: 19 Decimals: 31 Flags: BINARY +---------------------------------------------------------------------+ | date_sub(concat(substr(datetime,1,13),':00:00'), interval 1 second) | +---------------------------------------------------------------------+ | 2015-09-18 08:59:59 | +---------------------------------------------------------------------+ 1 row in set (0.00 sec) Is there some global variable I can set to revert to the 5.1 behavior? I realize I can CAST the result to a datetime to make sure I don't get the fractional seconds, but as I mentioned we're converting our system over to 5.5 and there's boatloads of code that I would have to scan through to try and find these differences. 3.) If there is no way to revert to the 5.1 behavior, can someone help me to understand the exact reason why the milliseconds get displayed on the final select in MariaDB 5.5? I'll need to look for any similar types of results and other than the exact reduction above I'm not sure what statements I would need to identify and test. Thanks for any assistance! Dan Ragle

3 7

[Maria-discuss] doubt about merge table (10000+) with same structure
by Roberto Spadim 23 Sep '15

23 Sep '15

Hi again guys :) i'm with a new problem i have MANY (10000+) tables with same struct, and i want to execute a select from "some" (100+) tables i was thinking about creating a view with all tables, something like: CREATE VIEW view_name AS SELECT * FROM table1 WHERE primary_key=<a value only at table1> UNION ALL SELECT * FROM table2 WHERE primary_key=<a value only at table2> UNION ALL SELECT * FROM table3 WHERE primary_key=<a value only at table3> UNION ALL SELECT * FROM table4 WHERE primary_key=<a value only at table4> UNION ALL ... SELECT * FROM tablen WHERE primary_key=<a value only at tablen> but i don't know if mysql optimizer will do a good job when i execute something like: SELECT * FROM view_name WHERE primary_key=<a value only located at table1> and just execute the query at table1 instead of alllllllll 10000+ tables i'm considering restruture database to a spider engine or any other method if i don't have other option any idea is well come -- Roberto Spadim

3 4

[Maria-discuss] WHY re TMM files never removed in 10.0.x
by Reindl Harald 20 Sep '15

20 Sep '15

i was flooded once again with mcron-mails "ERROR 144 (HY000) at line 1: Table './asterisk/cdr' is marked as crashed and last (automatic?) repair failed" well, because mariadb creates such temp files and don't remove them which makes myisam-recover-options = "FORCE" useless -rw-rw---- 1 mysql mysql 0 2015-08-21 13:47 cdr.TMM MariaDB [asterisk]> optimize table cdr; +--------------+----------+----------+------------------------------------------------------------------------+ | Table | Op | Msg_type | Msg_text | +--------------+----------+----------+------------------------------------------------------------------------+ | asterisk.cdr | optimize | error | Can't create new tempfile: '/Volumes/dune/mysql_data/asterisk/cdr.TMM' | | asterisk.cdr | optimize | status | Operation failed | +--------------+----------+----------+------------------------------------------------------------------------+

2 5

[Maria-discuss] Cross compile issue
by Johnson Cheng 18 Sep '15

18 Sep '15

Hi, I have problem to do cross compile on mariadb-10.0.21. Where can I find the relative document? Best Regards, Johnson

2 17

[Maria-discuss] New Question: Which is the best solution to re-sync data if one of my mariadb servers was offline for few days?
by AskMonty KB 18 Sep '15

18 Sep '15

Hello, A new question has been asked in "MariaDB Community" by long_beo_g. Please answer it at http://mariadb.com/kb/en/which-is-the-best-solution-to-re-sync-data-if-one-… as the person asking the question may not be subscribed to the mailing list. -------------------------------- Hi all, In our system, we built only two Mariadb Galera servers. We use keepalived module to load balancing IP address . But we found out one of two servers offline for few days. When I start the offline server (server A), the online one (server B) take a long time to replicate to server A, run out of the server's resources. Could Mariadb masters tell me how to deal with this issue ? How could I sync online data to offline one, and make two servers run well online ? Thank you. -------------------------------- To view or answer this question please visit: http://mariadb.com/kb/en/which-is-the-best-solution-to-re-sync-data-if-one-…

1 0

[Maria-discuss] ODBC.NET DOWNLOAD
by Bruce Carlson 17 Sep '15

17 Sep '15

Hi, I'm trying to download the latest stable version of the MariaDB .net ODBC install MSI but every time I try I get directed to a log in page where I can't get any further unless I create some sort of personal information divulgence account. I have never had to do this before with any open source software. Where can I go to download the MSI without having to create an account that I have no intentions of ever doing. I have been using MySQL ever since the early days of version 3 something and have been using MariaDB ever since inception after the split with MySQL. I'm a dedicated and heavy user and supporter of MariaDB and have been for decades. If I am no longer able to download updates etc. without having to create some sort of GOOGLE account then I will not use MariaDB, I will find an alternative. Please advise. Bruce Carlson

3 3

Re: [Maria-discuss] ODBC.NET DOWNLOAD
by Bruce Carlson 17 Sep '15

17 Sep '15

Hi Jan, Appreciate your comments however the issue is not the availability of the source or my ability to compile my own binaries from the source and I would even be glad to pay a fee for the convenience of downloading a pre built MSI when installing on windows machines and I am reasonably happy with the occasional donations to open source I've been able to extract from my clients. (Although, I admit, we could all possibly do more in the area of digging up donations and raising funds is a constant issue for all open source organisations). My complaint is I object to being SPECIFICALLY targeted by advertising. The only benefit derived from insisting on this type of "account" is to directly target advertising. THERE IS NO OTHER BENEFIT. I do not object to advertising. When I'm browsing MariaDB I expect to get advertising from MariaDB and maybe some affiliates. And MariaDB has some excellent affiliates. I'm on the mailing lists and forums of several open source communities. They constantly send me advertising matter. I contribute. I donate. I do OBJECT to DIRECT TARGETING. Here is just one example of how this works :- Every time I use a web browser to search for any subject, the search engine will target me with advertising from those businesses that have specified my credentials with the search provider thereby denying me access to alternatives. And that is a very real concern and in my opinion is totally contrary to the principles of open source. It's at least contrary to my principles. Now you may think this is trivial and I'm over reacting but I want to be the person in charge of what information I'm fed. No matter how trivial that sounds to you. In the end, if the only way I can get hold of the binaries is to download the source and build my own, I will or find an alternative product. On a completely separate note, Whilst I disagree with you on the subject of direct targeting I completely agree with your attitude to renewables. And I feel just as strongly about both subjects. Regards, Bruce Carlson -----Original Message----- From: Jan Steinman [mailto:Jan@EcoReality.org] Sent: Thursday, 17 September 2015 12:26 PM To: Bruce Carlson Cc: Rasmus Johansson; Maria Discuss (maria-discuss(a)lists.launchpad.net) Subject: Re: [Maria-discuss] ODBC.NET DOWNLOAD On 2015-09-16, at 18:13, Bruce Carlson wrote: > the openness of open SOURCE [emphasis mine] Do you think you may be over-reacting just a bit? My understanding -- correct me if I'm wrong -- is that the SOURCE is freely downloadable, and that the extra content you vociferously object to is only presented if you choose to download a BINARY. Compiling, maintaining, and debugging binaries is NOT part of the "open source" agreement! Rather, it is a service for which one might expect to be compensated. If you want to compile and maintain and debug the SOURCE and provide binaries without advertising, YOU CAN! NO ONE IS STOPPING YOU! Again, my apology if I'm misunderstanding something, but I used Homebrew to download and compile the SOURCE for MariaDB, and I don't recall seeing ANY advertising of any sort. :::: We are like tenant farmers chopping down the fence around our house for fuel when we should be using nature's inexhaustible sources of energy - sun, wind and tide. I'd put my money on the sun and solar energy. What a source of power! I hope we don't have to wait until oil and coal run out before we tackle that. -- Thomas Alva Edison :::: Jan Steinman, EcoReality Co-op ::::

1 0

[Maria-discuss] New Question: Galera mysqld_multi - possible bug
by AskMonty KB 16 Sep '15

16 Sep '15

Hello, A new question has been asked in "MariaDB Community" by cisaksen. Please answer it at http://mariadb.com/kb/en/galera-mysqld_multi-possible-bug/ as the person asking the question may not be subscribed to the mailing list. -------------------------------- Not sure where to post this but I've come across this on a SLES 12 system. Running Galera 25.3.9-1 / MariaDB 10.1.6-1. Using the Mysqld_multi init.d script to start and stop multiple mysql instances through service or systemd commands I have found that the starting works just fine but stopping them doesn't. It will stop all the instances but the last one. Just for clarification cmd is : service mysqld_multi stop or systemctl stop mysqld_multi. I tried this with 2 and 3 mysql instances. With 2 instances, #1 would stop but #2 wouldn't. With 3 instances #1 & #2 would stop and #3 wouldn't. I have verified that the mysqladmin command works to shut the DB down with the multi_admin account I use. Now running mysqld_multi directly works just fine. Whether it's one specific instance or all of them. That is running the script directly /etc/init.d/mysqd_multi or the actual /usr/bin/mysqld_multi command directly. Both work as expected. I honestly don't know if this is a script problem or a systemd problem. I do know that in SLES 12, SUSE rewrote the mysql init.d script to include mysqld_multi if multiple instances were discovered in the my.cnf file. And this work great using systemd. Anyway there it is, if anyone has come across this on other systems it might be worth noting. -------------------------------- To view or answer this question please visit: http://mariadb.com/kb/en/galera-mysqld_multi-possible-bug/

1 0