- developers - lists.mariadb.org

[Maria-developers] [GSOC 2014] CREATE OR REPLACE, CREATE IF NOT EXISTS, and DROP IF EXISTS
by Kesha Shah 03 Apr '14

03 Apr '14

Dear Sergei Golubchik, Hope this mail finds you well. I am Kesha Shah, a third year undergraduate from DA-IICT, Gandhinagar, India. I was a Google Summer of Code student previous year as well as Google Code-In mentor with BRL-CAD open source solid modeling software. The activities related to my contributions at BRL-CAD at logged by me at - http://brlcad.org/wiki/User:KeshaSShah/GSoC13/Reports. My other projects can be found at - https://github.com/keshashah?tab=repositories. I have good programming hands on C and C++ and would like to take up and contribute at the Mariadb task of 'CREATE OR REPLACE, CREATE IF NOT EXISTS, and DROP IF EXISTS' [MDEV-5359] as part of GSOC 2014. I have studied database systems and compiler construction during my academics. Your project would help to apply my knowledge by working at deeper level of coding database internals, thereby contributing to open source software. I had checked out the mariadb master branch from bazaar and started looking into the source code. I found the task more related to update of the existing mariadb SQL parser. The parser seems to be implemented as ascendant parser using bison parser generator (GNU version of YACC). From the current findings I could see the YACC grammar file sql_yacc.yy under maria/sql/ where I can see the rules defined for CREATE PROCEDURE_SYM, CREATE FUNCTION_SYM, CREATE TRIGGER_SYM etc. Alternatively I can see the *_SYM lexical tokens defined under the /maria/sql/lex.h file. Similarly for DROP TABLESPACE_SYM, DROP TRIGGER_SYM etc could be found. So when sql text input stream is encountered from SQL interface, it is read and categorized into lexical tokens by the lexer /maria/sql/sql_lex.cc in the lexical analysis step. I tried to figure out that we will have to define additional token 'IF' in the lex.h and 'NOT' and 'EXISTS' are already defined. I am also looking at the existing user defined actions that builds the AST nodes (sp_head being the root node for most types) when relevant grammar rule is satisfied in case of DDL statements. After semantic analysis (create of stored procs/functions), the next step would be to amend/define actions for the newly introduced grammar of handling check of existence of db artefact at the time of DDL execution. Also at this step the compiler will have to query the relevant storage engine on DB level or repository level to check if the DB object exists. This would require me to spend more time on understanding design pattern followed for the implementation of classes and behaviours of each of the underlying db artefacts. It would be interesting for me to know if the steps proposed by me above are in the right direction or not ? Also I am aware that I am late in discussing my proposal with you. But I can assure that I will be able to deliver the requirements in timely manner under your guidance. I have formulated initial draft of my proposal and would like to submit it at the earliest. Looking forward to you reply. Have a nice day! Thanks and Regards, Kesha Shah

2 2

[Maria-developers] new histogram code in 10.0.10
by Axel Schwenke 02 Apr '14

02 Apr '14

Hi Sergei, attached the results from a DBT3/SF30 run vs. 10.0.9 (as released) and 10.0-bzr rev 4128. Both runs use EITS + histograms. I see different EXPLAIN plans for queries 2, 3, 5, 7, 8, 18 and 21. Execution times only for few queries: 10.0.9 rev4128 Q2 44,44,46 44,45,45 Q3 298,298,299 236,243,243 Q5 264,302,303 293,308,308 Q7 199,200,201 204,205,206 Q8 375,375,377 264,274,276 Q18 217,218,221 178,180,182 Q21 251,259,261 258,261,262 Details in attached zip file. XL

1 0

[Maria-developers] PIC build options and MariaDB
by Otto Kekäläinen 02 Apr '14

02 Apr '14

Hello, I noticed that the MySQL packages in Debian have PIC stuff in their alternative build arguments, see http://anonscm.debian.org/gitweb/?p=pkg-mysql/mysql-5.6.git;a=blob;f=debian… http://anonscm.debian.org/gitweb/?p=pkg-mysql/mysql-5.5.git;a=blob;f=debian… Anybody here who understands deeply the build arguments who could say if having these PIC additions is desirable? Also the MySQL rules file has a section like this: # See http://blogs.innodb.com/wp/2010/04/innodb-performance-aio-linux/ 67 # Also #659565 68 ifneq (,$(findstring linux-gnu,$(DEB_HOST_GNU_TYPE))) 69 USE_LINUX_NATIVE_AIO:=-DLINUX_NATIVE_AIO=ON 70 endif Should MariaDB use this too, is it valid for XtraDB? -- Check out our blog at http://seravo.fi/blog and follow @ottokekalainen

2 1

[Maria-developers] Cassandra plugin build error: _S_empty_rep_storage' can not be used when making a shared object; recompile with -fPIC
by Otto Kekäläinen 02 Apr '14

02 Apr '14

Hello! I am testing out building MariaDB 10.0.10 on a Ubuntu 12.04 system with libthrift-dev installed (latest 0.9.0 from repo 'deb http://ppa.launchpad.net/wnoronha/thrift/ubuntu precise main'). Building fails with this error: make[4]: Siirrytään hakemistoon "/home/otto/MariaDB/pkg-mariadb-10.0/mariadb-10.0/builddir" Scanning dependencies of target cassandra make[4]: Poistutaan hakemistosta "/home/otto/MariaDB/pkg-mariadb-10.0/mariadb-10.0/builddir" make[4]: Siirrytään hakemistoon "/home/otto/MariaDB/pkg-mariadb-10.0/mariadb-10.0/builddir" [ 75%] Building CXX object storage/cassandra/CMakeFiles/cassandra.dir/ha_cassandra.cc.o [ 75%] Building CXX object storage/cassandra/CMakeFiles/cassandra.dir/cassandra_se.cc.o [ 75%] Building CXX object storage/cassandra/CMakeFiles/cassandra.dir/gen-cpp/Cassandra.cpp.o [ 75%] Building CXX object storage/cassandra/CMakeFiles/cassandra.dir/gen-cpp/cassandra_types.cpp.o [ 75%] Building CXX object storage/cassandra/CMakeFiles/cassandra.dir/gen-cpp/cassandra_constants.cpp.o Linking CXX shared module ha_cassandra.so /usr/bin/ld.bfd.real: /usr/lib/gcc/x86_64-linux-gnu/4.6/../../../../lib/libthrift.a(TApplicationException.o): relocation R_X86_64_32S against `std::basic_string<char, std::char_traits<char>, std::allocator<char> >::_Rep::_S_empty_rep_storage' can not be used when making a shared object; recompile with -fPIC /usr/lib/gcc/x86_64-linux-gnu/4.6/../../../../lib/libthrift.a: could not read symbols: Bad value collect2: ld:n paluuarvo oli 1 make[4]: *** [storage/cassandra/ha_cassandra.so] Virhe 1 Ideas? -- Check out our blog at http://seravo.fi/blog and follow @ottokekalainen

3 3

Re: [Maria-developers] Cassandra plugin build error: _S_empty_rep_storage' can not be used when making a shared object; recompile with -fPIC
by Otto Kekäläinen 01 Apr '14

01 Apr '14

Hello! Thanks for a quick reply! So I should add -fPIC to cmake arguments? Why isn't it in MariaDB sources by default, what are the downsides? GCC manual said "This option makes a difference on the m68k, PowerPC and SPARC." -> I am running amd64 For background, in my Debian packaging the build command looks like this: sh -c 'PATH=${MYSQL_BUILD_PATH:-"/bin:/usr/bin"} \ CC=${MYSQL_BUILD_CC:-gcc} \ CFLAGS=${MYSQL_BUILD_CFLAGS:-"-O2 -DBIG_JOINS=1 -fno-strict-aliasing "} \ CXX=${MYSQL_BUILD_CXX:-g++} \ CXXFLAGS=${MYSQL_BUILD_CXXFLAGS:-"-O3 -DBIG_JOINS=1 -felide-constructors -fno-exceptions -fno-rtti -fno-strict-aliasing "} \ cmake -DCMAKE_INSTALL_PREFIX=/usr \ \ -DCOMPILATION_COMMENT="(Ubuntu)" \ -DMYSQL_SERVER_SUFFIX="-1" \ -DSYSTEM_TYPE="debian-linux-gnu" \ -DWITHOUT_TOKUDB=true \ -DBUILD_CONFIG=mysql_release \ -DINSTALL_LIBDIR=lib/x86_64-linux-gnu \ -DINSTALL_PLUGINDIR=lib/mysql/plugin \ -DINSTALL_SUPPORTFILESDIR=share/mysql \ -DDEB=1 ..' In upstream MariaDB the build command is: sh -c 'PATH=$${MYSQL_BUILD_PATH:-"/bin:/usr/bin:/usr/local/bin"} \ CC=$${MYSQL_BUILD_CC:-gcc} \ CFLAGS=$${MYSQL_BUILD_CFLAGS:-"-O2 -fno-omit-frame-pointer -g -pipe -Wall -Wno-uninitialized $$(case `lsb_release -sc` in (lenny) echo -DWORKAROUND_GCC_4_3_2_BUG ;; esac)"} \ CXX=$${MYSQL_BUILD_CXX:-g++} \ CXXFLAGS=$${MYSQL_BUILD_CXXFLAGS:-"-O2 -fno-omit-frame-pointer -g -pipe -Wall -Wno-uninitialized"} \ cmake .. \ -DMYSQL_SERVER_SUFFIX="-$(DEBVERSION)" \ -DBUILD_CONFIG=mysql_release \ -DCOMPILATION_COMMENT="mariadb.org binary distribution" \ -DSYSTEM_TYPE="debian-linux-gnu" \ -DDEB=debian' 2014-04-01 16:21 GMT+03:00 Peter W Bowey <support(a)pbcomp.com.au>: > You require -fPIC to compile. See the info you reported = "> object; > recompile with -fPIC" > > Info is here (via GCC manual): > > -fPIC > If supported for the target machine, emit position-independent code, > suitable for dynamic linking and avoiding any limit on the size of the > global offset table. This option makes a difference on the m68k, PowerPC and > SPARC. > Position-independent code requires special support, and therefore works only > on certain machines. > > When this flag is set, the macros __pic__ and __PIC__ are defined to 2. > > In brief, the term position independent code (PIC) refers to the generated > machine code which is memory address agnostic, i.e. does not make any > assumptions about where it was loaded into RAM. Only position independent > code is supposed to be included into shared objects (SO) as they should have > an ability to dynamically change their location in RAM. > > > ----- Original Message ----- From: "Otto Kekäläinen" <otto(a)seravo.fi> > To: <maria-developers(a)lists.launchpad.net> > Cc: "Roger Meier" <roger(a)bufferoverflow.ch>; "Jake Farrell" > <jfarrell(a)apache.org> > Sent: Tuesday, April 01, 2014 11:16 PM > Subject: [Maria-developers] Cassandra plugin build error: > _S_empty_rep_storage' can not be used when making a shared object; recompile > with -fPIC > > >> Hello! >> >> I am testing out building MariaDB 10.0.10 on a Ubuntu 12.04 system >> with libthrift-dev installed (latest 0.9.0 from repo 'deb >> http://ppa.launchpad.net/wnoronha/thrift/ubuntu precise main'). >> >> Building fails with this error: >> >> make[4]: Siirrytään hakemistoon >> "/home/otto/MariaDB/pkg-mariadb-10.0/mariadb-10.0/builddir" >> Scanning dependencies of target cassandra >> make[4]: Poistutaan hakemistosta >> "/home/otto/MariaDB/pkg-mariadb-10.0/mariadb-10.0/builddir" >> make[4]: Siirrytään hakemistoon >> "/home/otto/MariaDB/pkg-mariadb-10.0/mariadb-10.0/builddir" >> [ 75%] Building CXX object >> storage/cassandra/CMakeFiles/cassandra.dir/ha_cassandra.cc.o >> [ 75%] Building CXX object >> storage/cassandra/CMakeFiles/cassandra.dir/cassandra_se.cc.o >> [ 75%] Building CXX object >> storage/cassandra/CMakeFiles/cassandra.dir/gen-cpp/Cassandra.cpp.o >> [ 75%] Building CXX object >> storage/cassandra/CMakeFiles/cassandra.dir/gen-cpp/cassandra_types.cpp.o >> [ 75%] Building CXX object >> >> storage/cassandra/CMakeFiles/cassandra.dir/gen-cpp/cassandra_constants.cpp.o >> Linking CXX shared module ha_cassandra.so >> /usr/bin/ld.bfd.real: >> >> /usr/lib/gcc/x86_64-linux-gnu/4.6/../../../../lib/libthrift.a(TApplicationException.o): >> relocation R_X86_64_32S against `std::basic_string<char, >> std::char_traits<char>, std::allocator<char> >>> >>> ::_Rep::_S_empty_rep_storage' can not be used when making a shared >> >> object; recompile with -fPIC >> /usr/lib/gcc/x86_64-linux-gnu/4.6/../../../../lib/libthrift.a: could >> not read symbols: Bad value >> collect2: ld:n paluuarvo oli 1 >> make[4]: *** [storage/cassandra/ha_cassandra.so] Virhe 1 >> >> >> Ideas? >> >> >> -- >> Check out our blog at http://seravo.fi/blog >> and follow @ottokekalainen >> >> _______________________________________________ >> Mailing list: https://launchpad.net/~maria-developers >> Post to : maria-developers(a)lists.launchpad.net >> Unsubscribe : https://launchpad.net/~maria-developers >> More help : https://help.launchpad.net/ListHelp >> > > > --- > This email is free from viruses and malware because avast! Antivirus > protection is active. > http://www.avast.com > -- Check out our blog at http://seravo.fi/blog and follow @ottokekalainen

1 0

[Maria-developers] DBT-3 benchmarks: MariaDB 10 vs MySQL 5.6
by Colin Charles 31 Mar '14

31 Mar '14

Hi XL, Nice work on posting new benchmarks in replication. Are you planning on running DBT-3 benchmarks with 10.0.10 and latest 5.6? Maybe even 5.7 DMR? Thanks -- Colin Charles, Chief Evangelist, SkySQL - The MariaDB Company blog: http://bytebot.net/blog/| t: +6-012-204-3201 | Skype: colincharles

1 0

[Maria-developers] maria-developers@lists.launchpad.net
by Sergey Vojtovich 31 Mar '14

31 Mar '14

Hi! This is second list of questions I came up while merging test cases: - --ignore-builtin-innodb is deprecated in 5.6, do we want to deprecate it too? - LOW_PRIORITY WRITE is deprecate in 5.6, do we want to deprecate it too? The following items seem to be not merged to MariaDB. Do we want to merge them? WL#5331 Support Unicode for Windows command line client Bug#12688860 : SECURITY RECOMMENDATION: PASSWORDS ON CLI WL#5855 (EXPLAIN FORMAT=JSON) partition test WL#5980 - Allow remote tablespaces (need to check) WL5602 Providing password hashing with appropriate strength Bug#12316645: Wrong cost calculation with optimizer_join_cache_level settings WL#5602 - SHA256 password authentication Bugs#12635232: VALGRIND WARNINGS: IS_IPV6, IS_IPV4, INET6_ATON, INET6_NTOA + MULTIBYTE CHARSET Bug#14040277 UNINITIALIZED VALUE REFERENCED IN STR_TO_IPV6 And a lot of other IPV6 related stuff. Regards, Sergey

2 2

[Maria-developers] Intermediate status for test cases merge
by Sergey Vojtovich 31 Mar '14

31 Mar '14

Hi! So far I processed 431 test cases from the main suite. Most of them are up to date, a few merged without problems, a few tests fail (bugs reported), a few need extra analysis. There are a few items which I'd like to discuss: - INSERT DELAYED is deprecated in 5.6, do we want to deprecate it too? - SHOW PROFILE is deprecated in 5.6, do we want to deprecate it too? - YEAR(2) is deprecated in 5.6, do we want to deprecate it too? Monty suggests that we shouldn't deprecate it. I find it Ok too. Reasons for YEAR(2) deprecation are not obvious, relevant worklog is private. Relevant revision comment says: "YEAR(2) is a subject to deprecation since it has ill design." - EXPLAIN gives NULL in Extra column in 5.6 (empty string in 10.0), do we want behave similarly? - Do we want to merge fix for BUG#60269 - mysql should reject attempts to create system tables in incorrect engine http://bugs.mysql.com/bug.php?id=60269 - NO_ENGINE_SUBSTITUTION is default mode in 5.6 now, do we want it be default too? - 5.6 has many updates to default values, do we like to merge them too? Thanks, Sergey

2 6

Re: [Maria-developers] [Commits] Rev 3995: MDEV-5138 Numerous test failures in "mtr --ps --embedded"
by Sergei Golubchik 31 Mar '14

31 Mar '14

Hi, Holyfoot! On Dec 09, holyfoot(a)askmonty.org wrote: > At file:///home/hf/wmar/mdev-5138/ > > ------------------------------------------------------------ > revno: 3995 > revision-id: holyfoot(a)askmonty.org-20131209020845-9o3l30ao90vzf600 > parent: holyfoot(a)askmonty.org-20131208143921-3veatgg9wngyzb47 > committer: Alexey Botchkov <holyfoot(a)askmonty.org> > branch nick: mdev-5138 > timestamp: Mon 2013-12-09 06:08:45 +0400 > message: > MDEV-5138 Numerous test failures in "mtr --ps --embedded". > If a prepared statement calls an stored procedure, > the thd->server_status out of the SP goes up > to the PS and then to the client. So that the > client gets the SERVER_STATUS_CURSOR_EXISTS status > if the SP uses a cursor. Which makes the embedded > server fail. Why does embedded fail while a normal client-server protocol is ok with that? > Fixed by saving/restoring the upper-level server_status > in sp_head::execute(). Regards, Sergei

2 3

[Maria-developers] Engine independent statistics - naming issues?
by Sergey Petrunia 31 Mar '14

31 Mar '14

Hello, I was looking at whether column names used by EIS feature are consistent with themselves and with ther parts of the server. == Database / table / column names == There are a few naming conventions arready in use: 1. Old school: mysql.columns_priv, tables_priv Db Table_name Column_name 2. SQL STANDARD ALL CAPS HELLO FROM 1970S: information_schema.* (tables, columns, indexes, index_statistics, etc) TABLE_CATALOG (always 'def') TABLE_SCHEMA TABLE_NAME COLUMN_NAME INDEX_NAME 3. InnoDB has its own convention: mysql.innodb_table_stats, innodb_index_stats database_name table_name index_name 4. EITS feature will add: mysql.column_stats, index_stats, table_stats db_name table_name column_name index_name == Other == There are variables @@histogram_size, @@histogram_type. mysql.column_stats has columns named hist_size, hist_type. What's your opinion, should we make some renames to make things more consistent? BR Sergei -- Sergei Petrunia, Software Developer MariaDB | Skype: sergefp | Blog: http://s.petrunia.net/blog

2 1