- developers - lists.mariadb.org

[Maria-developers] buildbot config
by Leif Walsh 09 Jul '13

09 Jul '13

Hi, I'm learning buildbot and I was wondering if I could take a look at the master.cfg for buildbot.askmonty.org as a reference for learning common practices and things like that. Is it available somewhere? -- Cheers, Leif

2 2

[Maria-developers] merge for MySQL56 FSP data types
by Alexander Barkov 09 Jul '13

09 Jul '13

Hi Serg, Please review the patch merging MySQL-5.6 fractional second precision enabled data types. Thanks.

3 7

[Maria-developers] Connect: compilation warnings
by Alexander Barkov 08 Jul '13

08 Jul '13

Hi Oliver, I fixed almost all compilation warnings that happened in Connect on my Fedora 17 laptop with gcc-4.7.2. Most of them were "variable is initialized but not used" and "no previous declaration". There were not serious, but very unpleasant for developers. (Note, we try to keep the tree in a shape that it compiles without warnings. Connect must follow this rule as well!!!) I think there was a bug in one place: how SQLSetStmtAttr() was called. See "bzr diff -c 3827" for details. The value of m_RowsetSize was erroneously passed, but SQLSetStmtAttr() expects the address of a SQLUINT variable (not a value!). See the ODBC docs for the details: http://msdn.microsoft.com/en-us/library/windows/desktop/ms712631%28v=vs.85%… One compilation warning still remains: maria-10.0-connect/storage/connect/tabsys.cpp:279:26: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast] It looks like a bug. Cast from 4-byte integer to 8-byte pointer is obviously wrong. Please address the problem. Thanks.

1 0

[Maria-developers] Warning in dynamic columns code
by Sergei Petrunia 08 Jul '13

08 Jul '13

Hi Sanja, I have been seeing this warning on a number of platforms: 10.0-monty/sql/item_strfunc.cc: In member function ‘virtual bool Item_func_dyncol_create::fix_fields(THD*, Item**)’: 10.0-monty/sql/item_strfunc.cc:3887: warning: suggest a space before ‘;’ or explicit braces around empty body in ‘for’ statement I assume it is harmless, but just in case, passing it over to you. Could you please check it and get rid of it? Thanks Sergei -- Sergei Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

2 1

[Maria-developers] SO_REUSEPORT and mariadb
by Roberto Spadim 08 Jul '13

08 Jul '13

hi guys, i was reading the SO_REUSEPORT of kernel 3.9, and i didn't found it in mariadb maybe we could use it? here are the SO_REUSEPORT topics: https://lwn.net/Articles/542629/ http://linux.dell.com/files/presentations/Linux_Plumbers_Conf_2010/Scaling_… here the important parts: ------------------ 1)The first of the traditional approaches is to have a single listener thread that accepts all incoming connections and then passes these off to other threads for processing. The problem with this approach is that the listening thread can become a bottleneck in extreme cases. In early discussions<http://thread.gmane.org/gmane.linux.network/102140/focus=102150> on SO_REUSEPORT, Tom noted that he was dealing with applications that accepted 40,000 connections per second. Given that sort of number, it's unsurprising to learn that Tom works at Google. 2)The second of the traditional approaches used by multithreaded servers operating on a single port is to have all of the threads (or processes) perform an accept() call on a single listening socket in a simple event loop of the form: while (1) { new_fd = accept(...); process_connection(new_fd); } ------------------ i didn -- Roberto Spadim

1 1

[Maria-developers] Replication
by Roberto Spadim 05 Jul '13

05 Jul '13

hi guys, i don't know if it's the better place to ask... maybe in user list is the right one, please tell me what's better... i'm using replication, and i need to mark all tables that don't replicate, it's very sad boring time consuming, etc etc... there's a easier way to mark tables? maybe alter table xxxx REPLICATION=ON/OFF if not, could we add this feature? thanks guys -- Roberto Spadim

1 0

[Maria-developers] Developer course
by Roberto Spadim 05 Jul '13

05 Jul '13

hi guys, there's a internal mysql/mariadb developer course? i'm studying source, but it's a bit difficult understand what's the best class/methods to use, maybe a course could help me developing and understanding better the source any one know about a course like it? i will focus on query rewrite (select.cc), optimizer (opt_range.cc/opt_*) and partition engine, others engine (myisam,heap,innodb,aria) is not my priority now -- Roberto Spadim

2 3

[Maria-developers] Review request : MDEV-573 (mysqlreport update)
by Jean Weisbuch 04 Jul '13

04 Jul '13

Hi, I have been doing modifications to the MariaDB mysqlreport version (which is shipped on the Debian/Ubuntu packages and can be found on the debian/additions/ directory), especially to add stats/informations about MariaDB specific functions but also to fix some bugs and limitations of the tool. As of today, the script works without throwing a warning and seems to output correct informations on MariaDB 5.5 (either with or without threadpool and binary log activated), MariaDB 5.5 + TokuDB, MySQL 5.1, MySQL 5.5 and it solves some small bugs and issues that could lead warning messages, invalid values returned or truncated values. The main additions are the AriaDB pagecache stats, binlog cache stats, subquery cache stats, TokuDB stats and Threadpool stats. There are also some "cosmetic" changes/fixes : the addition of values such as the % of disk tables used or the InnoDB read "From file" modified to "From disk" as of the documentation of the status variables states ; values such as the threads cache size or the table cache size that were limited to 3 chars are now limited to 5 chars ; the per seconds stats was truncated to 5 chars max on Questions and Query Cache sections (for example 153.6k/s was truncated to 153.6/s). The JIRA task : https://mariadb.atlassian.net/browse/MDEV-573 The GIT repository where i put the latest revisions (i only have been using it lately but i commented all the changes and bugfixes along the time on the JIRA task and could provide older revisions/diff if required) : https://github.com/jb-boin/mariadb/blob/5.5.30/debian/additions/mysqlreport I would like to be sure that the values/stats i added are really useful to monitor (and the calculations of these correct) and that the way informations are shown, named and ordered is clear enough ; for example, the way the test that determines if its a MariaDB or a MySQL server might not be the best possible (it checks if the Aria engine is available). There is still an upstream bug on the "Questions" and the "SELECT and Sort" section, the %Total values are not correctly calculated and lead to values > 100%, i havent got time to give it a look yet and i included in the example output below values hitting that bug. Here is an example of the output of the latest revision on a server having about all the possible functions enabled (values are mixed from different servers) : MariaDB 5.5.30-tokudb-7.0.1 uptime 70 9:4:47 Thu Jul 4 01:48:24 2013 __ Key _________________________________________________________________ Buffer used 651.41M of 1.00G %Used: 63.61 Current 675.43M %Usage: 65.96 Write hit 79.79% Read hit 99.99% __ Questions ___________________________________________________________ Total 2.57M 0.4/s -Unknown 39.92M 6.6/s %Total: 1550.4 DMS 39.78M 6.5/s 1545.3 Com_ 2.71M 0.4/s 105.09 COM_QUIT 444 0.0/s 0.02 Slow 4 s 285 0.0/s 0.01 %DMS: 0.00 Log: OFF DMS 39.78M 6.5/s 1545.3 UPDATE 39.76M 6.5/s 1544.5 99.95 SELECT 21.11k 0.0/s 0.82 0.05 REPLACE 0 0/s 0.00 0.00 DELETE 0 0/s 0.00 0.00 INSERT 0 0/s 0.00 0.00 Com_ 2.71M 0.4/s 105.09 show_proces 2.51M 0.4/s 97.33 admin_comma 152.51k 0.0/s 5.92 show_status 9.23k 0.0/s 0.36 __ Rows ________________________________________________________________ Rows 186.00G 240.3k/s Using idx 168.06G 217.1k/s %Index: 90.35 Rows/question 158.90 __ SELECT and Sort _____________________________________________________ Scan 42.33k 0.0/s %SELECT: 200.53 Range 0 0/s 0.00 Full join 19.27k 0.0/s 91.29 Range check 0 0/s 0.00 Full rng join 0 0/s 0.00 Sort scan 0 0/s Sort range 0 0/s Sort mrg pass 0 0/s __ Table Locks _________________________________________________________ Waited 8.86M 11.4/s %Total: 0.46 Immediate 1.93G 2.5k/s __ Tables ______________________________________________________________ Open 6843 of 8192 %Cache: 83.53 Opened 14.14k 0.0/s __ Connections _________________________________________________________ Max used 5403 of 22000 %Max: 24.56 Total 176.48M 228.0/s __ Created Temp ________________________________________________________ Disk table 7.76k 0.0/s %Disk: 14.55 Table 53.38k 0.0/s Size: 64.0M File 12 0.0/s __ Threads _____________________________________________________________ Running 274 of 2788 Created 181.68k 0.2/s Slow 0 0/s Threadpool 892 of 1200 %Used: 74.33 Running 458 of 1200 %Running: 38.17 Idle 434 of 1200 %Idle: 36.17 __ Aborted _____________________________________________________________ Clients 5.03k 0.0/s Connects 3.65k 0.0/s __ Bytes _______________________________________________________________ Sent 1.51T 1.9M/s Received 445.91G 576.0k/s __ InnoDB Buffer Pool __________________________________________________ Usage 25.97G of 26.00G %Usage: 99.89 Read hit 100.00% Pages Free 1.83k %Total: 0.11 Data 1.67M 98.23 %Drty: 0.07 Misc 28381 1.67 Latched 0 0.00 Reads 527.54G 681.5k/s From disk 100.00k 0.1/s %Disk: 0.00 Ahead Rnd 0 0/s Writes 5.21G 6.7k/s Flushes 17.35M 22.4/s Wait Free 9 0.0/s %Wait: 0.00 __ InnoDB Lock _________________________________________________________ Waits 36660726 47.4/s Current 0 Time acquiring Total 627441780 ms Average 171 ms Max 1537520 ms __ InnoDB Data, Pages, Rows ____________________________________________ Data Reads 512.26k 0.7/s Writes 181.92M 235.0/s fsync 2.07M 2.7/s Pending Reads 0 Writes 0 fsync 0 Pages Created 3.72M 4.8/s Read 512.18k 0.7/s Written 17.35M 22.4/s Rows Deleted 175.85M 227.2/s Inserted 200.46M 259.0/s Read 174.52G 225.4k/s Updated 224.40M 289.9/s __ Aria Pagecache ______________________________________________________ Buffer used 3.43M of 512.00M %Used: 0.67 Current 20.26M %Usage: 3.96 Write hit 47.09% Read hit 99.96% __ Subquery Cache ______________________________________________________ Hit ratio 99.91% Hits 38.63G 49.9k/s Miss 35.97M 46.5/s __ Binary Log Cache _____________________________________________________ Disk use Transactional 0.43% Non transactional 0.50% __ TokuDB ______________________________________________________________ Cachetable 4.00G of 4.00G %Usage: 99.99 Miss 1.83M 0.3/s Evictions 1.95M 0.3/s Thanks in advance, Jean Weisbuch

1 0

Re: [Maria-developers] Rev 3824: MDEV-4472 Audit plugin
by Sergei Golubchik 03 Jul '13

03 Jul '13

Hi, Holyfoot! Generally, it looks pretty good. One comment, see below about MARIADB_ONLY. On Jul 02, holyfoot(a)askmonty.org wrote: > revno: 3824 > revision-id: holyfoot(a)askmonty.org-20130702213841-xwmzadaotzxqhz2k > parent: sergii(a)pisem.net-20130701155424-lx0dvaocejx0cn8l > committer: Alexey Botchkov <holyfoot(a)askmonty.org> > branch nick: mdev-4472 > timestamp: Wed 2013-07-03 02:38:41 +0500 > message: > MDEV-4472 Audit plugin. > Plugin implemented in plugin/server_audit/server_audit.c. > Rotating log file support modified to be mysql-service. > === added directory 'plugin/server_audit' > === added file 'plugin/server_audit/CMakeLists.txt' > --- a/plugin/server_audit/CMakeLists.txt 1970-01-01 00:00:00 +0000 > +++ b/plugin/server_audit/CMakeLists.txt 2013-07-02 21:38:41 +0000 > @@ -0,0 +1,16 @@ > +# Copyright (C) 2012 Monty Program Ab. > +# > +# This program is free software; you can redistribute it and/or modify > +# it under the terms of the GNU General Public License as published by > +# the Free Software Foundation; version 2 of the License. > +# > +# This program is distributed in the hope that it will be useful, > +# but WITHOUT ANY WARRANTY; without even the implied warranty of > +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the > +# GNU General Public License for more details. > +# > +# You should have received a copy of the GNU General Public License > +# along with this program; if not, write to the Free Software > +# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA > + > +MYSQL_ADD_PLUGIN(server_audit server_audit.c MODULE_ONLY) I'd remove MODULE_ONLY, if the plugin can be complied as a static plugin. No need to add artificial restrictions. > === added file 'plugin/server_audit/server_audit.c' > --- a/plugin/server_audit/server_audit.c 1970-01-01 00:00:00 +0000 > +++ b/plugin/server_audit/server_audit.c 2013-07-02 21:38:41 +0000 ... > +#ifndef MARIADB_ONLY > +#define MYSQL_SERVICE_LOGGER_INCLUDED > +#endif /*MARIADB_ONLY*/ Is MARIADB_ONLY normally defined? As far as I can see, it's not. So in the packages this plugin won't have MARIADB_ONLY. That's ok in your tree, but let's think about how to solve this before pushing into the main tree. Regards, Sergei

2 1

[Maria-developers] merge for MySQL56 temporal literals
by Alexander Barkov 03 Jul '13

03 Jul '13

Hi Serg, Please review the patch merging MySQL-5.6 temporal literals. Thanks.

3 4