- discuss - lists.mariadb.org

[Maria-discuss] Single-threaded performance & optimizer overhead
by Colin Charles 03 Apr '13

03 Apr '13

Hi! Mark Callaghan asks this on Facebook: While testing single-threaded performance to compare MySQL 5.1 with 5.6 I think there is much more overhead in the optimizer in 5.6. Do you have benchmark results for a single-threaded workload with fast queries (like SELECT one row by PK from a cached table)? I am curious whether new versions of MariaDB have a similar increase in optimizer overhead. http://bugs.mysql.com/bug.php?id=68825 He also wrote a blog post: http://mysqlha.blogspot.com/2013/04/mysql-56-single-thread-update-only.html Might make a great post at our blog cheers, -colin -- Colin Charles, http://bytebot.net/blog/ | twitter: @bytebot | skype: colincharles MariaDB: Community developed. Feature enhanced. Backward compatible. Download it at: http://www.mariadb.org/ Open MariaDB/MySQL documentation at the Knowledgebase: http://kb.askmonty.org/

3 2

Re: [Maria-discuss] [Dbmail] MariaDB and dbmail
by Reindl Harald 03 Apr '13

03 Apr '13

Am 03.04.2013 11:09, schrieb Sergej Pupykin: > At Tue, 02 Apr 2013 19:47:15 +0400, > Sergej Pupykin <ml(a)sergej.pp.ru> wrote: >> Not sure if it caused by migration to mariadb, but problem appeared >> near same time. > > Downgrading back from mariadb 5.5.30 to mysql 5.5.30 solves the > problem. It looks like mariadb has ordering problems or it is not > drop-in replacement of mysql i am CC'ing this to the MariaDB list.......... but "however when I run it from command line it works well" speaks against it - this is highly interesting since more and more distributions are switching to MariaDB including Fedora and i am at preparing the switch too with no problems until now _____________________________ zamri <myzamri(a)gmail.com> wrote: > MariaDB is fully compatible with MySQL. Migration is done. So far so good. Phewww!... > > Now I need to do some research on tuning mariadb if needed. I found that after migration following SQL returns wrong sorted data: SELECT l.part_key, l.part_depth, l.part_order, l.is_header, DATE_FORMAT(ph.internal_date,GET_FORMAT(DATETIME,'ISO')), data FROM dbmail_mimeparts p JOIN dbmail_partlists l ON p.id = l.part_id JOIN dbmail_physmessage ph ON ph.id = l.physmessage_id WHERE l.physmessage_id = <ID> ORDER BY l.part_key,l.part_order ASC; however when I run it from command line it works well. Probably it does not fail always because of not all emails damaged. Not sure if it caused by migration to mariadb, but problem appeared near same time.

3 3

Re: [Maria-discuss] [Dbmail] MariaDB and dbmail
by Sergej Pupykin 03 Apr '13

03 Apr '13

At Wed, 03 Apr 2013 14:35:59 +0400, Sergej Pupykin <ml(a)sergej.pp.ru> wrote: > > Yes, but > > $ systemctl stop postfix dbmail-imapd dbmail-timsieved dbmail-pop3d mysqld > $ pacman -S mysql-clients mysql libmysqlclient > $ systemctl start mysqld postfix dbmail-imapd dbmail-timsieved dbmail-pop3d > $ mysql_upgrade --force > > solves the problem. > > I tested it using telnet to imap port. > ? LOGIN ... > ? SELECT ... > ? FETCH <ID> body[] > and check order of mime parts. > > Also dbmail/libzdb code looks clean. It just get sql query results and > put them to IMAP client... Link to thread start: http://www.mail-archive.com/dbmail%40dbmail.org/msg18179.html

1 0

[Maria-discuss] [Certification] Oracle MySQL exams horribly outdated. Are there MariaDB exams in planning?
by Lukas Lehner 26 Mar '13

26 Mar '13

Oracle MySQL exams are horribly outdated. Are there MariaDB exams in planning?

2 1

[Maria-discuss] WL#4925 and multi-source replication
by Greg 15 Mar '13

15 Mar '13

Hi, I'm testing multi-source replication with 6 sources, it works great ! Now I'm at the "HA" state, testing a master multi-source with DRBD. To do a safe switch between masters, I have to enable sync_binlog=1 and then the master delay grow... I see that WL#4925 suggest a solution that could work : is that job in progress in MariaDB 10 trunk ? And if not, how to ask to include it ? Thanks! -- Greg

4 12

[Maria-discuss] MariaDB 5.5.30 now available
by Daniel Bartholomew 12 Mar '13

12 Mar '13

The MariaDB project is pleased to announce the immediate availability of MariaDB 5.5.30. This is a bug-fix release. See the Release Notes and Changelog for details. - - Links - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - MariaDB 5.5.30 Stable (GA) - Release Notes: https://kb.askmonty.org/en/mariadb-5530-release-notes - Changelog: https://kb.askmonty.org/en/mariadb-5530-changelog - Downloads: https://downloads.mariadb.org/mariadb/5.5.30 Overview of MariaDB 5.5: https://kb.askmonty.org/en/what-is-mariadb-55 APT and YUM Repository Configuration Generator: https://downloads.mariadb.org/mariadb/repositories/ - - User Feedback plugin - - - - - - - - - - - - - - - - - - - - - - - MariaDB includes a User Feedback plugin. This plugin is disabled by default. If enabled, it submits basic, completely anonymous MariaDB usage information. This information is used by the developers to track trends in MariaDB usage to better guide development efforts. If you would like to help make MariaDB better, please add "feedback=ON" to your my.cnf (my.ini on Windows) file! See http://kb.askmonty.org/en/user-feedback-plugin for more information. - - Quality - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - The project always strives for quality, but in reality, nothing is perfect. Please take time to report any issues you encounter at: http://mariadb.org/jira - - Support MariaDB - - - - - - - - - - - - - - - - - - - - - - - - - - If you would like to contribute to MariaDB, please see our "contributing" and "donations" pages. We also have merchandise available in a cafepress store. - https://kb.askmonty.org/en/contributing - https://mariadb.org/en/donate/ - http://www.cafepress.com/mariadb We hope you enjoy MariaDB! -- MariaDB: An Enhanced Drop-in Replacement for MySQL Website - http://mariadb.org Twitter - http://twitter.com/mariadb Google+ - http://google.com/+mariadb Facebook - http://fb.com/MariaDB.dbms Knowledgebase - http://kb.askmonty.org Monty Program - http://montyprogram.com

1 0

[Maria-discuss] unexpected index growth
by Said Ramirez 11 Mar '13

11 Mar '13

We have noticed a very strange index growth with one of our tables. The data in the table is currently ~ 42GB and has been growing at a rate of about 1 GB per day for the week. Inserts/Updates on the tables are done with statements like: insert into foo ( ...) on duplicate key update ... We have a job which updates about 1/3 of the table 12 times a day. We running Mariadb version 5.5.27-MariaDB-log We noticed that the index sizes have been growing quite a bit: ** 7 days ago *** > select table_name,index_name,index_total_pages, (index_total_pages * 16*1024)/(1024*1024*1024) index_size_gb from INNODB_INDEX_STATS where table_name='foo'; +--------------------+----------------------+-------------------+---------------+ | table_name | index_name | index_total_pages | index_size_gb | +--------------------+----------------------+-------------------+---------------+ | foo | PRIMARY | 1192456 | 18.1954 | | foo | idx_active_deleted | 5647809 | 86.1787 | | foo | idx_last_update_plus | 5329538 | 81.3223 | | foo | active | 8586719 | 131.0229 | +--------------------+----------------------+-------------------+---------------+ 4 rows in set (0.00 sec) > optimize table db.foo; +---------+----------+----------+-------------------------------------------------------------------+ | Table | Op | Msg_type | Msg_text | +---------+----------+----------+-------------------------------------------------------------------+ | db.foo | optimize | note | Table does not support optimize, doing recreate + analyze instead | | db.foo | optimize | status | OK | +---------------------------------+----------+----------+-------------------------------------------------------------------+ 2 rows in set (4 hours 1 min 24.73 sec) > select table_name,index_name,index_total_pages, (index_total_pages * 16*1024)/(1024*1024*1024) index_size_gb from INNODB_INDEX_STATS where table_name='foo'; +--------------------+----------------------+-------------------+---------------+ | table_name | index_name | index_total_pages | index_size_gb | +--------------------+----------------------+-------------------+---------------+ | foo | PRIMARY | 639296 | 9.7549 | | foo | idx_active_deleted | 403520 | 6.1572 | | foo | idx_last_update_plus | 531392 | 8.1084 | | foo | active | 517312 | 7.8936 | +--------------------+----------------------+-------------------+---------------+ 4 rows in set (0.00 sec) ** Today ** > select table_name,index_name,index_total_pages, (index_total_pages * 16*1024)/(1024*1024*1024) index_size_gb from INNODB_INDEX_STATS where table_name='foo'; +--------------------+----------------------+-------------------+---------------+ | table_name | index_name | index_total_pages | index_size_gb | +--------------------+----------------------+-------------------+---------------+ | foo | PRIMARY | 1013436 | 15.4638 | | foo | idx_active_deleted | 2478512 | 37.8191 | | foo | idx_last_update_plus | 2459700 | 37.5320 | | foo | active | 3888175 | 59.3288 | +--------------------+----------------------+-------------------+---------------+ 4 rows in set (0.10 sec) One of the even stranger things is that the slaves of this table do not experience such a noticeable index size growth: [slave] >select table_name,index_name,index_total_pages, (index_total_pages * 16*1024)/(1024*1024*1024) index_size_gb from INNODB_INDEX_STATS where table_name='foo'; +--------------------+----------------------+-------------------+---------------+ | table_name | index_name | index_total_pages | index_size_gb | +--------------------+----------------------+-------------------+---------------+ | foo | PRIMARY | 1010492 | 15.4189 | | foo | idx_active_deleted | 454510 | 6.9353 | | foo | idx_last_update_plus | 543085 | 8.2868 | | foo | active | 583468 | 8.9030 | +--------------------+----------------------+-------------------+---------------+ 4 rows in set (0.01 sec) Our suspicion is that it has to do with how transactions are executing on the master since on the slaves the insert/updates would be executed in a single thread. However we can't tell why this is happening. Is this a known bug we are hitting ? As anyone else seen this behavior and come up with a solution other than having to run optimize table ? Thanks, -Said

2 4

[Maria-discuss] MariaDB Galera Cluster 5.5.29 Stable (GA) Released
by Daniel Bartholomew 05 Mar '13

05 Mar '13

After some final testing and polishing, the MariaDB project and Codership are pleased to announce the release of MariaDB Galera Cluster 5.5.29. This is a Stable (GA) release. - - MariaDB Galera Cluster 5.5.29 Links - - - - - - - - - - - - - - - - Release Notes: http://kb.askmonty.org/en/mariadb-galera-5529-release-notes Changelog: http://kb.askmonty.org/en/mariadb-galera-5529-changelog Downloads: http://downloads.mariadb.org/mariadb-galera/5.5.29 APT & YUM Repository Configuration Tool: http://downloads.mariadb.org/mariadb-galera/repositories Overview of MariaDB Galera Cluster 5.5: http://kb.askmonty.org/en/what-is-mariadb-galera-cluster More information: http://kb.askmonty.org/en/galera http://codership.com - - About MariaDB Galera Cluster - - - - - - - - - - - - - - - - - - - MariaDB Galera Cluster is made for today's cloud based environments. It is fully read-write scalable, comes with synchronous replication, allows multi-master topologies, and guarantees no lag or lost transactions. Some of its features & benefits are listed below. Features * Synchronous replication * Active-active multi-master topology * Read and write to any cluster node * Automatic membership control, with failed nodes dropped from the cluster * Automatic node joining * True row-level parallel replication * Direct client connections, native MariaDB/MySQL look & feel Benefits The above features yield several benefits for a database clustering solution, including: * No slave lag * No lost transactions * Both read and write scalability * Smaller client latencies Technology MariaDB Galera Cluster uses the Galera Library for the replication implementation. http://codership.com/products/galera_replication To interface with Galera replication, we have enhanced MariaDB to support the replication API definition in the Write Set REPlication (https://launchpad.net/wsrep) project. The implementation of the replication API in MariaDB happens in the open source MySQL-wsrep project. http://www.codership.com/products/mysql-write-set-replication-project - - Operations Support - - - - - - - - - - - - - - - - - - - - - - - - Once you run MariaDB Galera Cluster in production you want to make sure there is a helping hand available whenever needed. We're happy to inform you that our partner, SkySQL, offers 24-hour support for MariaDB Galera Cluster, see http://www.skysql.com/enterprise/galera-cluster for more information. - - User Feedback plugin - - - - - - - - - - - - - - - - - - - - - - - MariaDB Galera Cluster includes a User Feedback plugin. This plugin is disabled by default. If enabled, it submits basic, completely anonymous usage information. This information is used by the developers to track trends in usage to better guide development efforts. If you would like to help make MariaDB Galera Cluster better, please add "feedback=ON" to your my.cnf file! See http://kb.askmonty.org/en/user-feedback-plugin for more information. - - Quality - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - The project always strives for quality, but in reality, nothing is perfect. Please take time to report any issues you encounter at: http://mariadb.org/jira We hope you enjoy MariaDB Galera Cluster! - - Where else to find us - - - - - - - - - - - - - - - - - - - - - - - http://mariadb.org We're also on: Google+: http://google.com/+mariadb Facebook: http://fb.com/MariaDB.dbms Twitter: http://twitter.com/mariadb Follow, friend, and add us to your circles to keep up to date with MariaDB! -- MariaDB: An Enhanced Drop-in Replacement for MySQL

1 0

[Maria-discuss] KB Question: FreeBSD 9.1, Mariadb55, and strfill()
by Daniel Bartholomew 04 Mar '13

04 Mar '13

The following question has been posted to the Knowledgebase: https://kb.askmonty.org/en/freebsd-91-mariadb55-and-strfill/ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - FreeBSD 9.1, Mariadb55, and strfill() Please give me a hint. Under FreeBSD my application cannot find strfill(). It can be found in libmysqlclient.a and libmysqlclient_r.a but NOT in the shared library libmysqlclient.so. Oddly, strcend() can be found in all of those libraries but I cannot find a difference in CMake between the two. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Thanks. -- Daniel Bartholomew Google+ - http://gplus.to/dbart Twitter - http://twitter.com/daniel_bart MariaDB: An Enhanced Drop-in Replacement for MySQL Website - http://mariadb.org Twitter - http://twitter.com/mariadb Google+ - http://google.com/+mariadb Facebook - http://fb.com/MariaDB.dbms Knowledgebase - http://kb.askmonty.org Monty Program - http://montyprogram.com

1 0

[Maria-discuss] Using 10.0s multi-source replication for database consolidation
by Markus "Shorty" Uckelmann 22 Feb '13

22 Feb '13

Hi all, I want to consolidate the databases on several servers onto a single server. And all this without or only minimal downtime. After reading an article[1] from the Percona folks I was thinking of using the multi-source replication feature of 10.0. But there are a few things I'm not sure about. Maybe someone can help me out. What MySQL/MariaDB versions are needed before and behind the MariaDB 10 server? My plan/idea/dream was the following: 1. Stop the server on a slave and start it with skip-slave and skip-networking. Write down replication data. 2. Dump the databases I want to move with "mysqldump <db-name>" 3. Load the dump into MariaDB 10.0 and configure replication. This is done for all the databases I want to put onto a single server. The rest should be business as usual. BTW: We use similar grants on all servers. So the MariaDB 10.0 will be setup with the same statements and I don't have to care about the "mysql"-database. Does this sound realistic? Or have I missed something? [1] http://www.mysqlperformanceblog.com/2012/12/18/migrating-several-single-sta… Thanks a lot, Shorty

2 4