- developers - lists.mariadb.org

[Maria-developers] Gsoc 2016 Mdev 371 Unique index for blob
by Sachin Setia 16 Mar '16

16 Mar '16

Dear Developers, I was doing prototype for this project as Mr Sergei Golubchik suggested "This task is about creating MI_UNIQUEDEF "uniques" instead of MI_KEYDEF "keys" for long unique constraints." Here is my output MariaDB [sachin]> create table tbl(int_key int primary key , blob_key blob unique); Query OK, 0 rows affected (0.03 sec) MariaDB [sachin]> insert into tbl values(1,1); Query OK, 1 row affected (0.01 sec) MariaDB [sachin]> insert into tbl values(2,1); ERROR 1062 (23000): Duplicate entry '1' for key 'blob_key' MariaDB [sachin]> insert into tbl values(2,2); Query OK, 1 row affected (0.00 sec) MariaDB [sachin]> insert into tbl values(3,3); Query OK, 1 row affected (0.00 sec) MariaDB [sachin]> select * from tbl; +---------+----------+ | int_key | blob_key | +---------+----------+ | 1 | 1 | | 2 | 2 | | 3 | 3 | +---------+----------+ 3 rows in set (0.00 sec) I am assuming that unique blob will have 0 zero key . Should i mail you the source code.Please let me know Regards sachin

2 4

[Maria-developers] Introduction - GSOF16
by Chandu Maddila 16 Mar '16

16 Mar '16

Hi Team, I'm Chandra Sekhar Maddila from India and currently masters student of BITS-Pilani. I'm pretty interested in working with MariaDB team as part of GSOF. Some of the items that I'm interested in are MXS-1, MDEV-8100, MDEV-5662, MDEV-371 Here's a quick summary of my work experience and Skill set. - Currently a student of BITS-Pilani <http://www.bits-pilani.ac.in/> (Masters, Computer Science). - Total experience of 6 years in IT industry, as a software developer (My last stint was in *Microsoft Research *[MSR] - India as a research SDE). - Have been working with designing and development of RDBMS systems : With Oracle earlier and SQL Server after joining Microsoft - Involved in various projects like :* Implementing column store indexes in SQL Server, Designing a cloud based monitoring systems for database performance monitoring, tools to migrate data from Oracle to SQL Server, Implementing HA Solution (AlwaysOn) for SQL Server etc...* - Languages/Frameworks: - Asp.Net, C#, Jquery, Javascript, HTML5, CSS, Bootstrap, Knockout, SOAP, REST, JSON, Web Services - WCF, MVVM, XAML, Windows App Development, Azure Cloud services, Azure web services - Java, J2EE, UNIX Shell scripting I worked extensively on both *UNIX and Windows platforms*. *Database domain expertise:* *Oracle* + In depth understanding of Oracle database internals and architecture (Memory management, storage engine and optimizer) + *Performance tuning*: Instance and SQL level tuning with the help of *AWR, Statspack, TKPROF* reports and Explain plans and tools like *OEM* and *TOAD* + Oracle *11g, 10g, 9i* Database Setup, Administration, Monitoring on various flavors of UNIX: RHEL, AIX, SUN Solaris, HPUX – Itanium chipset; Good understanding of Oracle internals and Oracle system architecture + Oracle *RAC, ASM and Grid infrastructure* setup, maintenance and administration; Non RAC to RAC, Non-ASM to ASM conversions; Familiarity with OS level clustering techniques like *HACMP* + All kinds of Database *Backup and Recovery* methodologies: *RMAN *(Hot, Cold, Active Database Duplication)*, Datapump, SQL Loader* etc…; Familiarity with 3rd party backup tools like *CA ArcServe* *SQL Server* + Working experience with SQL 2014, 2012, 2008R2, 2008 and 2005 on different versions of Windows + Very good at *troubleshooting complex issues and working on outages/showstoppers*. + Very good in *database performance and query tuning*: Analyzing *execution plans* and rewriting queries + Administering SQL Server High Availability and Disaster Recovery technologies like *AlwaysOn, Failover Clustering *and *Database mirroring*. + Proficiency and expertise in all varieties of *SQL Server Replication*: *Snapshot Replication, Transactional replication, Merge replication and Change tracking* + Expertise in handling all sorts of *backup technologies: Native, VSS, VDI, Backup to URL etc…* + Expert in handling *database corruption cases and recovering* the databases + Expertise in T-SQL writing stored procedures, triggers and functions. For any queries, feel free to reach out to me @ chandu.maddila(a)gmail.com Thanks, Chandra

2 1

Re: [Maria-developers] Gsoc 2016 Mdev 371 Unique index for blob
by Sergei Golubchik 15 Mar '16

15 Mar '16

Hi, Sachin! First: please send questions like this to maria-developers(a)lists.launchpad.net it's a public mailing list dedicated to MySQL and MariaDB internals, source code, and related things. I am subscribed, so I'll see you mail there, and you may be sure I will, because it won't be accidentally catched by my spam filter, or sorted out in some obscure folder. Furthermore other subscribers will see your question and could reply if I will be not available (e.g. I could be travelling). Thank you. On Mar 14, Sachin Setia wrote: > Thank you sir for your reply but my question is what this pack flag will do > and basically what is pack record will you please suggest source file in > which i can look into to get more understanding of pack length ,pack flag > and pack record There isn't much I can answer - your question is too generic. "pack_flag" is a set of flags that is stored per field in the frm file and it describes some aspects of the field. Search for "pack_flag" in the source tree. If you ask a specific question, I'll be able to provide a more detailed answer. Regards, Sergei Chief Architect MariaDB and security(a)mariadb.org

1 0

Re: [Maria-developers] [Commits] a0c06ba: Preliminary implementation for the aggregate sum function as a window function
by Sergey Petrunia 14 Mar '16

14 Mar '16

Hi Vicentiu, I think it add_helper() is a really poor choice of name when the function can do removal. Maybe add_or_remove() name would be better? On Mon, Mar 14, 2016 at 03:44:48PM +0200, Vicentiu Ciorbaru wrote: > revision-id: a0c06ba1edb54c8c4705189c0455137a85658297 (mariadb-10.1.8-154-ga0c06ba) > parent(s): ce8a0d8e19e7bf1486a2c80ff6b7a30ef35bf99f > author: Vicențiu Ciorbaru > committer: Vicențiu Ciorbaru > timestamp: 2016-03-14 15:42:00 +0200 > message: > > Preliminary implementation for the aggregate sum function as a window function > > This implementation does not deal with the case where removal of > elements from the window frame causes the item to turn to a null value. > > --- > mysql-test/r/win_sum.result | 42 ++++++++++++++++++++++++++++++++++++++++++ > mysql-test/t/win_sum.test | 32 ++++++++++++++++++++++++++++++++ > sql/item_sum.cc | 29 +++++++++++++++++++++++++---- > sql/item_sum.h | 4 ++++ > sql/sql_window.cc | 1 + > 5 files changed, 104 insertions(+), 4 deletions(-) > > diff --git a/mysql-test/r/win_sum.result b/mysql-test/r/win_sum.result > new file mode 100644 > index 0000000..1db6c6e > --- /dev/null > +++ b/mysql-test/r/win_sum.result > @@ -0,0 +1,42 @@ > +create table t1 ( > +pk int primary key, > +a int, > +b int, > +c real > +); > +insert into t1 values > +(101 , 0, 10, 1.1), > +(102 , 0, 10, 2.1), > +(103 , 1, 10, 3.1), > +(104 , 1, 10, 4.1), > +(108 , 2, 10, 5.1), > +(105 , 2, 20, 6.1), > +(106 , 2, 20, 7.1), > +(107 , 2, 20, 8.15), > +(109 , 4, 20, 9.15), > +(110 , 4, 20, 10.15), > +(111 , 5, NULL, 11.15), > +(112 , 5, 1, 12.25), > +(113 , 5, NULL, 13.35), > +(114 , 5, NULL, 14.50), > +(115 , 5, NULL, 15.65); > +select pk, a, b, sum(b) over (partition by a order by pk ROWS BETWEEN 1 PRECEDING AND 1 FOLLOWING), > +sum(c) over (partition by a order by pk ROWS BETWEEN 1 PRECEDING AND 1 FOLLOWING) > +from t1; > +pk a b sum(b) over (partition by a order by pk ROWS BETWEEN 1 PRECEDING AND 1 FOLLOWING) sum(c) over (partition by a order by pk ROWS BETWEEN 1 PRECEDING AND 1 FOLLOWING) > +101 0 10 20 3.2 > +102 0 10 20 3.2 > +103 1 10 20 7.199999999999999 > +104 1 10 20 7.199999999999999 > +105 2 20 40 13.2 > +106 2 20 60 21.35 > +107 2 20 50 20.35 > +108 2 10 30 13.250000000000002 > +109 4 20 40 19.3 > +110 4 20 40 19.3 > +111 5 NULL 1 23.4 > +112 5 1 1 36.75 > +113 5 NULL 1 40.1 > +114 5 NULL 0 43.5 > +115 5 NULL 0 30.15 > +drop table t1; > diff --git a/mysql-test/t/win_sum.test b/mysql-test/t/win_sum.test > new file mode 100644 > index 0000000..3c12b08 > --- /dev/null > +++ b/mysql-test/t/win_sum.test > @@ -0,0 +1,32 @@ > +create table t1 ( > + pk int primary key, > + a int, > + b int, > + c real > +); > + > + > +insert into t1 values > +(101 , 0, 10, 1.1), > +(102 , 0, 10, 2.1), > +(103 , 1, 10, 3.1), > +(104 , 1, 10, 4.1), > +(108 , 2, 10, 5.1), > +(105 , 2, 20, 6.1), > +(106 , 2, 20, 7.1), > +(107 , 2, 20, 8.15), > +(109 , 4, 20, 9.15), > +(110 , 4, 20, 10.15), > +(111 , 5, NULL, 11.15), > +(112 , 5, 1, 12.25), > +(113 , 5, NULL, 13.35), > +(114 , 5, NULL, 14.50), > +(115 , 5, NULL, 15.65); > + > +--sorted_result > +select pk, a, b, sum(b) over (partition by a order by pk ROWS BETWEEN 1 PRECEDING AND 1 FOLLOWING), > + sum(c) over (partition by a order by pk ROWS BETWEEN 1 PRECEDING AND 1 FOLLOWING) > + > +from t1; > + > +drop table t1; > diff --git a/sql/item_sum.cc b/sql/item_sum.cc > index c78c206..3f4853a 100644 > --- a/sql/item_sum.cc > +++ b/sql/item_sum.cc > @@ -1318,25 +1318,39 @@ void Item_sum_sum::fix_length_and_dec() > bool Item_sum_sum::add() > { > DBUG_ENTER("Item_sum_sum::add"); > + add_helper(false); > + DBUG_RETURN(0); > +} > + > +void Item_sum_sum::add_helper(bool perform_removal) > +{ > + DBUG_ENTER("Item_sum_sum::add_helper"); > if (Item_sum_sum::result_type() == DECIMAL_RESULT) > { > my_decimal value; > const my_decimal *val= aggr->arg_val_decimal(&value); > if (!aggr->arg_is_null(true)) > { > - my_decimal_add(E_DEC_FATAL_ERROR, dec_buffs + (curr_dec_buff^1), > - val, dec_buffs + curr_dec_buff); > + if (perform_removal) > + my_decimal_sub(E_DEC_FATAL_ERROR, dec_buffs + (curr_dec_buff ^ 1), > + dec_buffs + curr_dec_buff, val); > + else > + my_decimal_add(E_DEC_FATAL_ERROR, dec_buffs + (curr_dec_buff ^ 1), > + val, dec_buffs + curr_dec_buff); > curr_dec_buff^= 1; > null_value= 0; > } > } > else > { > - sum+= aggr->arg_val_real(); > + if (perform_removal) > + sum-= aggr->arg_val_real(); > + else > + sum+= aggr->arg_val_real(); > if (!aggr->arg_is_null(true)) > null_value= 0; > } > - DBUG_RETURN(0); > + DBUG_VOID_RETURN; > } > > > @@ -1386,6 +1400,13 @@ my_decimal *Item_sum_sum::val_decimal(my_decimal *val) > return val_decimal_from_real(val); > } > > +void Item_sum_sum::remove() > +{ > + DBUG_ENTER("Item_sum_sum::remove"); > + add_helper(true); > + DBUG_VOID_RETURN; > +} > + > /** > Aggregate a distinct row from the distinct hash table. > > diff --git a/sql/item_sum.h b/sql/item_sum.h > index f1f5b22..24dfd9b 100644 > --- a/sql/item_sum.h > +++ b/sql/item_sum.h > @@ -773,6 +773,10 @@ class Item_sum_sum :public Item_sum_num, > return has_with_distinct() ? "sum(distinct " : "sum("; > } > Item *copy_or_same(THD* thd); > + void remove(); > + > +private: > + void add_helper(bool perform_removal); > }; > > > diff --git a/sql/sql_window.cc b/sql/sql_window.cc > index df9899e..f0d74b1 100644 > --- a/sql/sql_window.cc > +++ b/sql/sql_window.cc > @@ -1584,6 +1584,7 @@ bool JOIN::process_window_functions(List<Item> *curr_fields_list) > } > case Item_sum::COUNT_FUNC: > case Item_sum::SUM_BIT_FUNC: > + case Item_sum::SUM_FUNC: > { > /* > Frame-aware window function computation. It does one pass, but > _______________________________________________ > commits mailing list > commits(a)mariadb.org > https://lists.askmonty.org/cgi-bin/mailman/listinfo/commits -- BR Sergei -- Sergei Petrunia, Software Developer MariaDB Corporation | Skype: sergefp | Blog: http://s.petrunia.net/blog

2 1

[Maria-developers] Gsoc 2016 Mdev 371 Unique index for blob
by Sachin Setia 14 Mar '16

14 Mar '16

Hello Developers My name is sachin , a third year undergrad student in India.I am interested in gsoc 2016 I was debugging the maria db for following query create table c4(a1 int primary key,b1 int unique,c1 blob) engine=myisam; In mysql_prepare_create_table there is a function name prepare_create_field which sets the sql_field pack flag . for a1 field it sets the pack flag =27 . do not quiet get what it means.

2 1

Re: [Maria-developers] c0b1148: MDEV-9659 : AWS KMS encryption plugin
by Sergei Golubchik 12 Mar '16

12 Mar '16

Hi, Wlad! Just a couple of comments: On Mar 11, wlad(a)mariadb.com wrote: > + # Relax AWS C++ SDK unreasonably high requirements for CMake version. Use replace utility (from MariaDB build) > + # to patch their CMakeLists.txt Nice solution. But you might need to add a dependency, to ensure that this external project target is built after replace. Unless cmake can, somehow, derive it automatically (you can test it by building just this specific target in a clean tree). > + MY_DIR *dirp = my_dir(".", MYF(0)); > + if (!dirp) > + { > + sql_print_error("Can't scan current directory"); > + DBUG_RETURN(-1); > + } > + for (unsigned int i=0; i < dirp->number_of_files; i++) > + { > + > + KEY_INFO info; > + if (extract_id_and_version(dirp->dir_entry[i].name, &info.key_id, &info.key_version) == 0) > + { > + key_info_cache[KEY_ID_AND_VERSION(info.key_id, info.key_version)]= info; > + latest_version_cache[info.key_id]= max(info.key_version, latest_version_cache[info.key_id]); > + } > + } > + DBUG_RETURN(0); my_dir() uses malloc(), don't forget my_dirend() to free the memory. Regards, Sergei

2 1

Re: [Maria-developers] 66cd0f2: MDEV-9704: ALTER TABLE does not work from client
by Sergei Golubchik 12 Mar '16

12 Mar '16

Hi, OleksandrByelkin! On Mar 11, OleksandrByelkin wrote: > revision-id: 66cd0f2f02bb45707b38e983186ab8e7aae21da9 (mariadb-10.1.8-156-g66cd0f2) > parent(s): 75d8544c0bc611e82078b9e5f59caef9105e4892 > committer: Oleksandr Byelkin > timestamp: 2016-03-11 16:55:57 +0100 > message: > > MDEV-9704: ALTER TABLE does not work from client > > Do not try to send progress packets if client was not informed about them. > > --- > sql/sql_acl.cc | 2 -- > 1 file changed, 2 deletions(-) > > diff --git a/sql/sql_acl.cc b/sql/sql_acl.cc > index cd44673..3ae25d5 100644 > --- a/sql/sql_acl.cc > +++ b/sql/sql_acl.cc > @@ -11900,8 +11900,6 @@ static ulong parse_client_handshake_packet(MPVIO_EXT *mpvio, > (((ulonglong)uint4korr(net->read_pos + 28)) << 32); > client_capabilities|= ext_client_capabilities; > } > - else if (client_capabilities & CLIENT_PROGRESS_OBSOLETE) > - client_capabilities|= MARIADB_CLIENT_PROGRESS; Okay. And I think that now you don't need CLIENT_PROGRESS_OBSOLETE either. Regards, Sergei Chief Architect MariaDB and security(a)mariadb.org

1 0

Re: [Maria-developers] MDEV-9659 Encryption plugin that uses AWS Key management service
by Sergei Golubchik 12 Mar '16

12 Mar '16

Hi, Wlad! On Mar 03, wlad(a)mariadb.com wrote: > revision-id: c1ed3c0267da2ece7664e9f876314979114b2186 (mariadb-10.1.11-26-gc1ed3c0) > parent(s): a62db8186e40a62366bc86f49a0195e9b687d86d > author: Vladislav Vaintroub > committer: Vladislav Vaintroub > timestamp: 2016-03-03 21:29:29 +0100 > message: > > MDEV-9659 Encryption plugin that uses AWS Key management service Looks pretty good, thanks! See comment below... > diff --git a/cmake/plugin.cmake b/cmake/plugin.cmake > index c24239c..361cb59 100644 > --- a/cmake/plugin.cmake > +++ b/cmake/plugin.cmake > @@ -74,7 +74,9 @@ MACRO(MYSQL_ADD_PLUGIN) > SET(compat "with${compat}") > ENDIF() > > - IF (compat STREQUAL ".") > + IF (ARG_DISABLED) > + SET(howtobuild NO) > + ELSEIF (compat STREQUAL ".") > SET(howtobuild DYNAMIC) > ELSEIF (compat STREQUAL "with.") > IF (NOT ARG_MODULE_ONLY) > @@ -122,7 +124,7 @@ MACRO(MYSQL_ADD_PLUGIN) > > # Build either static library or module > IF (PLUGIN_${plugin} MATCHES "(STATIC|AUTO|YES)" AND NOT ARG_MODULE_ONLY > - AND NOT ARG_DISABLED AND NOT ARG_CLIENT) > + AND NOT ARG_CLIENT) > > IF(CMAKE_GENERATOR MATCHES "Makefiles|Ninja") > # If there is a shared library from previous shared build, > @@ -178,8 +180,7 @@ MACRO(MYSQL_ADD_PLUGIN) > SET (mysql_optional_plugins ${mysql_optional_plugins} PARENT_SCOPE) > ENDIF() > ELSEIF(PLUGIN_${plugin} MATCHES "(DYNAMIC|AUTO|YES)" > - AND NOT ARG_STATIC_ONLY AND NOT WITHOUT_DYNAMIC_PLUGINS > - AND NOT ARG_DISABLED) > + AND NOT ARG_STATIC_ONLY AND NOT WITHOUT_DYNAMIC_PLUGINS) > > ADD_VERSION_INFO(${target} MODULE SOURCES) > ADD_LIBRARY(${target} MODULE ${SOURCES}) a bit more of the explanation would've been nice. in the changeset comment or - even better - pushing this change separately. you can select what files to include in the commit, you know... > diff --git a/plugin/aws_key_management/CMakeLists.txt b/plugin/aws_key_management/CMakeLists.txt > new file mode 100644 > index 0000000..ec9ee46 > --- /dev/null > +++ b/plugin/aws_key_management/CMakeLists.txt > @@ -0,0 +1,144 @@ > +# We build parts of AWS C++ SDK as CMake external project > +# The restrictions of the SDK (https://github.com/awslabs/aws-sdk-cpp/blob/master/README.md) > +# are > + > +# - OS : Windows,Linux or OSX > +# - C++11 compiler : VS2013+, gcc 4.7+, clang 3.3+ > +# - libcurl development package needs to be present on Unixes > +# > +# Since we're using ExternalProject that reads from git repository > +# we also require GIT to be present on the build machine > + > + > +# Give message why the building this plugin is skipped (only if -DVERBOSE is defined) > +# and and bail out. > +MACRO(SKIP_AWS_PLUGIN msg) > + IF(DEFINED VERBOSE) VERBOSE - is it used somewhere else? Or you've invented it on the spot for this plugin? I understand that it can be used in other cmake files, but is it? > + MESSAGE(STATUS "Skip aws_key_management - ${msg}") > + RETURN() > + ENDIF() > +ENDMACRO() Ok, as discussed on irc, I'd suggest to add here a check whether AWS SDK is already installed. Like, CheckLibraryExists or CheckSymbolExists. And if not - then try to download and compile it (which would require new cmake, git, curl, etc). > +IF(CMAKE_VERSION VERSION_LESS "2.8.8") > + SKIP_AWS_PLUGIN("CMake is too old") > +ENDIF() > + > +FIND_PACKAGE(Git) > +IF(NOT GIT_FOUND) > + SKIP_AWS_PLUGIN("no GIT") > +ENDIF() > + > +INCLUDE(ExternalProject) > +IF (NOT(WIN32 OR APPLE OR (CMAKE_SYSTEM_NAME MATCHES "Linux"))) > + SKIP_AWS_PLUGIN("OS unsupported by AWS SDK") > +ENDIF() > + > +IF(UNIX) > + FIND_PACKAGE(CURL) > + IF(NOT CURL_FOUND) > + SKIP_AWS_PLUGIN("AWS C++ SDK requires libcurl development package") > + ENDIF() > + SET(PIC_FLAG -fPIC) > +ENDIF() > + > +SET(CXX11_FLAGS) > +SET(OLD_COMPILER_MSG "AWS SDK requires c++11 -capable compiler (minimal supported versions are g++ 4.7, clang 3.3, VS2103)") > +IF(CMAKE_CXX_COMPILER_ID MATCHES "GNU") > + EXECUTE_PROCESS(COMMAND ${CMAKE_CXX_COMPILER} -dumpversion OUTPUT_VARIABLE GCC_VERSION) eh? I thought cmake knows compiler versions... let me see This one: CMAKE_<LANG>_COMPILER_VERSION > + IF (GCC_VERSION VERSION_LESS 4.7) > + SKIP_AWS_PLUGIN("${OLD_COMPILER_MSG}") > + ENDIF() > + SET(CXX11_FLAGS "-std=c++11") Instead of checking the compiler version, I would rather suggest to check for a feature that you need to use. Like MY_CHECK_CXX_COMPILER_FLAG(-std=c++11) IF (NOT HAVE_CXX__STD_C__11) SKIP_AWS_PLUGIN("${OLD_COMPILER_MSG}") ENDIF() Something like that. Not sure about HAVE_CXX__STD_C__11 name, though. > +ELSEIF (CMAKE_CXX_COMPILER_ID MATCHES "Clang") > + IF (CMAKE_CXX_COMPILER_VERSION VERSION_LESS 3.3) > + SKIP_AWS_PLUGIN("${OLD_COMPILER_MSG}") > + ENDIF() > + SET(CXX11_FLAGS "-stdlib=libc++") Same. > +ELSEIF(MSVC) > + IF (MSVC_VERSION LESS 1800) > + SKIP_AWS_PLUGIN("${OLD_COMPILER_MSG}") > + ENDIF() > +ELSE() > + SKIP_AWS_PLUGIN("Compiler not supported by AWS C++ SDK") > +ENDIF() > + > +IF(WIN32) > + SET(EXTRA_SDK_CMAKE_FLAGS -DCMAKE_CXX_FLAGS_DEBUGOPT="" -DCMAKE_EXE_LINKER_FLAGS_DEBUGOPT="") For consistency you can put this inside ELSEIF(MSVC), that's where you set options for gcc and clang. > +ENDIF() > + > +FIND_PROGRAM(PERL_EXECUTABLE perl) > + > +IF(PERL_EXECUTABLE) > + # CMake minimimal version for building AWS SDK can be less than 3.1.2 > + # Downgrade this requirement. > + SET(AWS_SDK_PATCH_COMMAND > + PATCH_COMMAND ${PERL_EXECUTABLE} -i.bak -pe s/3.1.2/2.8/g ${CMAKE_BINARY_DIR}/aws-sdk-cpp/CMakeLists.txt) 1. Without perl it'll require 3.1.2, right? 2. Couldn't you do that with cmake alone? 3. Is that safe at all - you pull the latest version from git, they can start using cmake-3.1.2 features any time. > +ENDIF() > + > +ExternalProject_Add( > + aws_sdk_cpp > + GIT_REPOSITORY "https://github.com/awslabs/aws-sdk-cpp.git" > + UPDATE_COMMAND "" > + ${AWS_SDK_PATCH_COMMAND} > + SOURCE_DIR "${CMAKE_BINARY_DIR}/aws-sdk-cpp" > + CMAKE_ARGS > + -DBUILD_ONLY=aws-cpp-sdk-kms -DSTATIC_LINKING=1 > + "-DCMAKE_CXX_FLAGS_DEBUG=${CMAKE_CXX_FLAGS_DEBUG} ${PIC_FLAG}" > + "-DCMAKE_CXX_FLAGS_RELWITHDEBINFO=${CMAKE_CXX_FLAGS_RELWITHDEBINFO} ${PIC_FLAG}" > + "-DCMAKE_CXX_FLAGS_RELEASE=${CMAKE_CXX_FLAGS_RELEASE} ${PIC_FLAG}" > + "-DCMAKE_CXX_FLAGS_MINSIZEREL=${CMAKE_CXX_FLAGS_MINSIZEREL} ${PIC_FLAG}" > + ${EXTRA_SDK_CMAKE_FLAGS} > + -DCMAKE_INSTALL_PREFIX=${CMAKE_BINARY_DIR}/aws_sdk_cpp > + TEST_COMMAND "" > +) > + > +# Figure out where AWS installs SDK libraries > +# The below is defined in AWS SDK's CMakeLists.txt > +# (and their handling is weird, every OS has special install directory) > +if(WIN32) > + SET(SDK_INSTALL_BINARY_PREFIX "windows") > +ELSEIF(APPLE) > + SET(SDK_INSTALL_BINARY_PREFIX "mac") > +ELSEIF(UNIX) > + SET(SDK_INSTALL_BINARY_PREFIX "linux") > +ENDIF() > +IF(NOT APPLE) > + IF(CMAKE_SIZEOF_VOID_P EQUAL 8) > + SET(SDK_INSTALL_BINARY_PREFIX "${SDK_INSTALL_BINARY_PREFIX}/intel64") > + ELSE() > + SET(SDK_INSTALL_BINARY_PREFIX "${SDK_INSTALL_BINARY_PREFIX}/ia32") > + ENDIF() > +ENDIF() > +IF(CMAKE_CONFIGURATION_TYPES) > + SET(SDK_INSTALL_BINARY_PREFIX "${SDK_INSTALL_BINARY_PREFIX}/${CMAKE_CFG_INTDIR}") > +ENDIF() > + > +# We do not need to build the whole SDK , just 2 of its libs > +set(AWS_SDK_LIBS aws-cpp-sdk-core aws-cpp-sdk-kms) > + > +FOREACH(lib ${AWS_SDK_LIBS}) > + ADD_LIBRARY(${lib} STATIC IMPORTED) > + ADD_DEPENDENCIES(${lib} aws_sdk_cpp) > + SET(loc "${CMAKE_BINARY_DIR}/aws_sdk_cpp/lib/${SDK_INSTALL_BINARY_PREFIX}/${CMAKE_STATIC_LIBRARY_PREFIX}${lib}${CMAKE_STATIC_LIBRARY_SUFFIX}") > + SET_TARGET_PROPERTIES(${lib} PROPERTIES IMPORTED_LOCATION ${loc}) > + IF(WIN32) > + SET_TARGET_PROPERTIES(${lib} PROPERTIES IMPORTED_LINK_INTERFACE_LIBRARIES "bcrypt;winhttp;wininet;userenv") > + ELSE() > + SET_TARGET_PROPERTIES(${lib} PROPERTIES IMPORTED_LINK_INTERFACE_LIBRARIES "${SSL_LIBRARIES};${CURL_LIBRARIES}") > + ENDIF() > +ENDFOREACH() > + > +IF(CMAKE_SYSTEM_NAME MATCHES "Linux") > + # Need whole-archive , otherwise static libraries are not linked > + SET(AWS_SDK_LIBS -Wl,--whole-archive ${AWS_SDK_LIBS} -Wl,--no-whole-archive) > +ENDIF() > + > +SET_TARGET_PROPERTIES(aws_sdk_cpp PROPERTIES EXCLUDE_FROM_ALL TRUE) > +INCLUDE_DIRECTORIES(${CMAKE_BINARY_DIR}/aws_sdk_cpp/include) > +SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} ${CXX11_FLAGS}") > +MYSQL_ADD_PLUGIN(aws_key_management aws_key_management_plugin.cc > + MODULE_ONLY COMPONENT aws-key-management > + LINK_LIBRARIES ${AWS_SDK_LIBS} DISABLED) > diff --git a/plugin/aws_key_management/aws_key_management_plugin.cc b/plugin/aws_key_management/aws_key_management_plugin.cc > new file mode 100644 > index 0000000..fa7d042 > --- /dev/null > +++ b/plugin/aws_key_management/aws_key_management_plugin.cc > @@ -0,0 +1,611 @@ > +/* > + Copyright (c) 2016 MariaDB Corporation > + > + 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 */ > + > + > +#include <my_global.h> > +#include <my_pthread.h> > +#include <my_sys.h> > +#include <mysql/plugin_encryption.h> > +#include <my_crypt.h> > +#include <string.h> > +#include <stdio.h> > +#include <stdlib.h> > +#include <mysqld_error.h> > +#include <base64.h> > +#include <map> > +#include <algorithm> > +#include <string> > +#include <vector> > +#include <iterator> > +#include <sstream> > +#include <fstream> > +#ifndef _WIN32 > +#include <sys/types.h> > +#include <dirent.h> > +#endif Use cmake checks when there are some, HAVE_DIRENT_H, HAVE_SYS_TYPES_H, HAVE_STRING_H, HAVE_STDLIB_H. Hmm, HAVE_STDLIB_H doesn't make sense to me :) could you remove it from configure.cmake, please? In a separate changeset. May be HAVE_SYS_TYPES_H too, not sure. But do use HAVE_DIRENT_H and HAVE_STDLIB_H above instead of _WIN32. > +#include <aws/core/client/AWSError.h> > +#include <aws/core/utils/logging/AWSLogging.h> > +#include <aws/core/utils/logging/ConsoleLogSystem.h> > +#include <aws/kms/KMSClient.h> > +#include <aws/kms/model/DecryptRequest.h> > +#include <aws/kms/model/DecryptResult.h> > +#include <aws/kms/model/GenerateDataKeyWithoutPlaintextRequest.h> > +#include <aws/kms/model/GenerateDataKeyWithoutPlaintextResult.h> > +#include <aws/core/utils/Outcome.h> > +using namespace std; > +using namespace Aws::KMS; > +using namespace Aws::KMS::Model; > +using namespace Aws::Utils::Logging; > +extern void sql_print_error(const char *format, ...); > +extern void sql_print_warning(const char *format, ...); > +extern void sql_print_information(const char *format, ...); > + > + > +/* Plaintext key info struct */ > +struct KEY_INFO > +{ > + unsigned int key_id; > + unsigned int key_version; > + unsigned int length; > + unsigned char data[MY_AES_MAX_KEY_LENGTH]; > + bool load_failed; /* if true, do not attempt to reload?*/ > +public: > + KEY_INFO() : key_id(0), key_version(0), length(0), load_failed(false){}; > +}; > +#define KEY_ID_AND_VERSION(key_id,version) ((longlong)key_id << 32 | version) > + > +/* Cache for the latest version, per key id */ > +static std::map<uint, uint> latest_version_cache; > + > +/* Cache for plaintext keys */ > +static std::map<ulonglong, KEY_INFO> key_info_cache; > + > +static const char *key_spec_names[]={ "AES_128", "AES_256", 0 }; > + > +/* Plugin variables */ > +static char* master_key_id; > +static unsigned long key_spec; > +static unsigned long log_level; > +static int rotate_key; > + > +/* AWS functionality*/ > +static int aws_decrypt_key(const char *path, KEY_INFO *info); > +static int aws_generate_datakey(uint key_id, uint version); > + > +static int extract_id_and_version(const char *name, uint *id, uint *ver); > +static unsigned int get_latest_key_version(unsigned int key_id); > +static unsigned int get_latest_key_version_nolock(unsigned int key_id); > +static int load_key(KEY_INFO *info); > + > +/* Mutex to serialize access to caches */ > +static pthread_mutex_t mtx; Use PSI, please. That's mysql_mutex_t, etc. > + > +static Aws::KMS::KMSClient *client; > + > +/* Redirect AWS trace to error log */ > +class MySQLLogSystem : public Aws::Utils::Logging::FormattedLogSystem > +{ > +public: > + > + using Base = FormattedLogSystem; > + MySQLLogSystem(LogLevel logLevel) : > + Base(logLevel) > + { > + } > + virtual LogLevel GetLogLevel(void) const override > + { > + return (LogLevel)log_level; > + } > + virtual ~MySQLLogSystem() > + { > + } > + > +protected: > + virtual void ProcessFormattedStatement(Aws::String&& statement) override > + { > +#ifdef _WIN32 > + /* > + On Windows, we can't use C runtime functions to write to stdout, > + because we compile with static C runtime, so plugin has a stdout > + different from server. Thus we're using WriteFile(). > + */ > + DWORD nSize= (DWORD)statement.size(); > + DWORD nWritten; > + const char *s= statement.c_str(); > + HANDLE h= GetStdHandle(STD_OUTPUT_HANDLE); > + > + WriteFile(h, s, nSize, &nWritten, NULL); > +#else > + printf("%s", statement.c_str()); printf? You have declared sql_print_information above, why wouldn't you use it here? > +#endif > + } > +}; > + > +/* > + Read filenames from the current directory. > + This is used in plugin initilization. We check for > + plugin specific names aws-kms-key.<id>.<version> > + to find out what keys/key versions are already there. > +*/ > +static vector<string> read_dir() > +{ > + vector<string> v; > + DBUG_ENTER("read_dir"); > +#ifdef _WIN32 > + WIN32_FIND_DATA ffd; > + HANDLE h= FindFirstFile("*", &ffd); > + if (h == INVALID_HANDLE_VALUE) > + { > + DBUG_RETURN(v); > + } > + do > + { > + /* handle file possibly updates caches*/ > + v.push_back(ffd.cFileName); > + } while (FindNextFile(h, &ffd)); > + FindClose(h); > +#else > + DIR *dir= opendir("."); > + if (!dir) > + { > + DBUG_RETURN(v); > + } > + struct dirent *dp; > + while ((dp= readdir(dir)) != NULL) > + { > + v.push_back(dp->d_name); > + } > + closedir(dir); > +#endif > + DBUG_RETURN(v); ok. although you could've used my_dir and save yourself some troubles :) > +} > + > + > +/* > + Plugin initialization. > + > + Create KMS client and scan datadir to find out which keys and versions > + are present. > +*/ > +static int plugin_init(void *p) > +{ > + DBUG_ENTER("plugin_init"); > + client = new KMSClient(); > + if (!client) > + { > + sql_print_error("Can not initialize KMS client"); > + DBUG_RETURN(-1); > + } > + InitializeAWSLogging(Aws::MakeShared<MySQLLogSystem>("aws_key_management_plugin", (Aws::Utils::Logging::LogLevel) log_level)); > + pthread_mutex_init(&mtx, NULL); > + vector<string> filenames= read_dir(); > + for (size_t i= 0; i < filenames.size(); i++) > + { > + KEY_INFO info; > + if (extract_id_and_version(filenames[i].c_str(), &info.key_id, &info.key_version) == 0) > + { > + key_info_cache[KEY_ID_AND_VERSION(info.key_id, info.key_version)]= info; > + latest_version_cache[info.key_id]= max(info.key_version, latest_version_cache[info.key_id]); > + } > + } > + DBUG_RETURN(0); > +} > + > + > + > +static int plugin_deinit(void *p) > +{ > + DBUG_ENTER("plugin_deinit"); > + latest_version_cache.clear(); > + key_info_cache.clear(); > + pthread_mutex_destroy(&mtx); > + delete client; > + ShutdownAWSLogging(); > + DBUG_RETURN(0); > +} > + > +/* Generate filename to store the ciphered key */ > +static void format_keyfile_name(char *buf, size_t size, uint key_id, uint version) > +{ > + snprintf(buf, size, "aws-kms-key.%u.%u", key_id, version); > +} > + > +/* Extract key id and version from file name */ > +static int extract_id_and_version(const char *name, uint *id, uint *ver) > +{ > + int len; > + int n= sscanf(name, "aws-kms-key.%u.%u%n", id, ver, &len); > + if (n == 2 && *id > 0 && *ver > 0 && len == (int)strlen(name)) > + return 0; > + return 1; > +} > + > +/* > + Decrypt key stored in aws-kms-key.<id>.<version> > + Cache the decrypted key. > +*/ > +static int load_key(KEY_INFO *info) > +{ > + int ret; > + char path[256]; > + DBUG_ENTER("load_key"); > + DBUG_PRINT("enter", ("id=%u,ver=%u", info->key_id, info->key_version)); > + format_keyfile_name(path, sizeof(path), info->key_id, info->key_version); > + ret= aws_decrypt_key(path, info); > + if (ret) > + info->load_failed= true; > + > + latest_version_cache[info->key_id]= max(latest_version_cache[info->key_id], info->key_version); > + key_info_cache[KEY_ID_AND_VERSION(info->key_id, info->key_version)]= *info; > + > + if (!ret) > + { > + sql_print_information("AWS KMS plugin: loaded key %u, version %u, key length %u bit", > + info->key_id, info->key_version,(uint)info->length*8); > + } > + else > + { > + sql_print_warning("AWS KMS plugin: key %u, version %u could not be decrypted", > + info->key_id, info->key_version); > + } > + DBUG_RETURN(ret); > +} > + > + > +/* > + Get latest version for the key. > + > + If key is not decrypted yet, this function also decrypt the key > + and error will be returned if decryption fails. > + > + The reason for that is that InnoDB crashes > + in case errors are returned by get_key(), Shall we rather fix that in InnoDB? Do you remember where it crashes? > + > + A new key will be created if it does not exist, provided there is > + valid master_key_id. > +*/ > +static unsigned int get_latest_key_version(unsigned int key_id) > +{ > + unsigned int ret; > + DBUG_ENTER("get_latest_key_version"); > + > + pthread_mutex_lock(&mtx); > + ret= get_latest_key_version_nolock(key_id); why do you need get_latest_key_version_nolock if you never use it directly? > + pthread_mutex_unlock(&mtx); > + > + DBUG_PRINT("info", ("key=%u,ret=%u", key_id, ret)); > + DBUG_RETURN(ret); > +} > + > +static unsigned int get_latest_key_version_nolock(unsigned int key_id) > +{ > + KEY_INFO info; > + uint ver; > + DBUG_ENTER("get_latest_key_version_nolock"); > + ver= latest_version_cache[key_id]; > + if (ver > 0) > + { > + info= key_info_cache[KEY_ID_AND_VERSION(key_id, ver)]; > + } > + if (info.load_failed) > + { > + /* Decryption failed previously, don't retry */ > + DBUG_RETURN(ENCRYPTION_KEY_VERSION_INVALID); > + } > + else if (ver > 0) > + { > + /* Key encrypted already, return it*/ may be s/encrypted already/exists/ ? > + if (info.length > 0) > + DBUG_RETURN(ver); > + } > + else // (ver == 0) > + { > + /* Generate a new key, version 1 */ > + if (!master_key_id[0]) > + { > + my_printf_error(ER_UNKNOWN_ERROR, > + "Can't generate key %u, because aws_key_management_master_key_id parameter is not set", > + MYF(ME_JUST_WARNING), key_id); > + DBUG_RETURN(ENCRYPTION_KEY_VERSION_INVALID); > + } > + if (aws_generate_datakey(key_id, 1) != 0) > + DBUG_RETURN(ENCRYPTION_KEY_VERSION_INVALID); > + info.key_id= key_id; > + info.key_version= 1; > + info.length= 0; > + } > + > + if (load_key(&info)) > + DBUG_RETURN(ENCRYPTION_KEY_VERSION_INVALID); > + DBUG_RETURN(info.key_version); > +} > + > + > +/* > + Decrypt a file with KMS > +*/ > +static int aws_decrypt_key(const char *path, KEY_INFO *info) > +{ > + DBUG_ENTER("aws_decrypt_key"); > + > + /* Read file content into memory */ > + ifstream ifs(path, ios::binary | ios::ate); > + if (!ifs.good()) > + { > + sql_print_error("can't open file %s", path); > + } > + size_t pos = (size_t)ifs.tellg(); > + std::vector<char> contents(pos); > + ifs.seekg(0, ios::beg); > + ifs.read(&contents[0], pos); > + > + /* Decrypt data the with AWS */ > + DecryptRequest request; > + Aws::Utils::ByteBuffer byteBuffer((unsigned char *)contents.data(), pos); > + request.SetCiphertextBlob(byteBuffer); > + DecryptOutcome outcome = client->Decrypt(request); > + if (!outcome.IsSuccess()) > + { > + sql_print_error("AWS KMS plugin: Decrypt failed for %s : %s", path, > + outcome.GetError().GetMessage().c_str()); > + DBUG_RETURN(-1); > + } > + Aws::Utils::ByteBuffer plaintext = outcome.GetResult().GetPlaintext(); > + size_t len = plaintext.GetLength(); > + > + if (len > (int)sizeof(info->data)) > + { > + sql_print_error("AWS KMS plugin: encoding key too large for %s", path); > + DBUG_RETURN(ENCRYPTION_KEY_BUFFER_TOO_SMALL); > + } > + memcpy(info->data, plaintext.GetUnderlyingData(), len); > + info->length= len; > + DBUG_RETURN(0); > +} > + > + > +/* Generate a new datakey and store it a file */ > +static int aws_generate_datakey(uint keyid, uint version) > +{ > + > + DBUG_ENTER("aws_generate_datakey"); > + GenerateDataKeyWithoutPlaintextRequest request; > + request.SetKeyId(master_key_id); > + request.SetKeySpec(DataKeySpecMapper::GetDataKeySpecForName(key_spec_names[key_spec])); > + > + GenerateDataKeyWithoutPlaintextOutcome outcome; > + outcome= client->GenerateDataKeyWithoutPlaintext(request); > + if (!outcome.IsSuccess()) > + { > + sql_print_error("AWS KMS plugin : GenerateDataKeyWithoutPlaintext failed : %s - %s", > + outcome.GetError().GetExceptionName().c_str(), > + outcome.GetError().GetMessage().c_str()); > + DBUG_RETURN(-1); > + } > + > + string out; > + char filename[20]; > + Aws::Utils::ByteBuffer byteBuffer = outcome.GetResult().GetCiphertextBlob(); > + > + format_keyfile_name(filename, sizeof(filename), keyid, version); > + int fd= my_open(filename, O_RDWR | O_CREAT, 0); > + if (fd < 0) > + { > + sql_print_error("AWS KMS plugin: Can't create file %s", filename); > + DBUG_RETURN(-1); > + } > + size_t len= byteBuffer.GetLength(); > + if (my_write(fd, byteBuffer.GetUnderlyingData(), len, 0) != len) > + { > + sql_print_error("AWS KMS plugin: can't write to %s", filename); > + my_close(fd, 0); > + my_delete(filename, 0); > + DBUG_RETURN(-1); > + } > + my_close(fd, 0); > + sql_print_information("AWS KMS plugin: generated encrypted datakey for key id=%u, version=%u", > + keyid, version); > + DBUG_RETURN(0); > +} > + > +/* Key rotation for a single key */ > +static int rotate_single_key(uint key_id) > +{ > + uint ver; > + ver= latest_version_cache[key_id]; > + > + if (!ver) > + { > + my_printf_error(ER_UNKNOWN_ERROR, "key %u does not exist", MYF(ME_JUST_WARNING), key_id); > + return -1; > + } > + else if (aws_generate_datakey(key_id, ver + 1)) > + { > + my_printf_error(ER_UNKNOWN_ERROR, "Could not generate datakey for key id= %u, ver= %u", > + MYF(ME_JUST_WARNING), key_id, ver); > + return -1; > + } > + else > + { > + KEY_INFO info; > + info.key_id= key_id; > + info.key_version = ver + 1; > + if (load_key(&info)) > + { > + my_printf_error(ER_UNKNOWN_ERROR, "Could not load datakey for key id= %u, ver= %u", > + MYF(ME_JUST_WARNING), key_id, ver); > + return -1; > + } > + } > + return 0; > +} > + > +/* Key rotation for all key ids */ > +static int rotate_all_keys() > +{ > + int ret= 0; > + for (map<uint, uint>::iterator it= latest_version_cache.begin(); it != latest_version_cache.end(); it++) > + { > + ret= rotate_single_key(it->first); > + if (ret) > + break; > + } > + return ret; > +} > + > +static void update_rotate(MYSQL_THD, struct st_mysql_sys_var *, void *, const void *val) > +{ > + if (!master_key_id[0]) > + { > + my_printf_error(ER_UNKNOWN_ERROR, > + "aws_key_management_master_key_id must be set to generate new data keys", MYF(ME_JUST_WARNING)); > + return; > + } > + pthread_mutex_lock(&mtx); > + rotate_key= *(int *)val; > + switch (rotate_key) > + { > + case 0: > + break; > + case -1: > + rotate_all_keys(); > + break; > + default: > + rotate_single_key(rotate_key); > + break; > + } > + rotate_key= 0; > + pthread_mutex_unlock(&mtx); > +} > + > +static unsigned int get_key( > + unsigned int key_id, > + unsigned int version, > + unsigned char* dstbuf, > + unsigned int* buflen) > +{ > + KEY_INFO info; > + > + DBUG_ENTER("get_key"); > + pthread_mutex_lock(&mtx); > + info= key_info_cache[KEY_ID_AND_VERSION(key_id, version)]; > + if (info.length == 0 && !info.load_failed) > + { > + info.key_id= key_id; > + info.key_version= version; > + load_key(&info); > + } > + pthread_mutex_unlock(&mtx); > + > + if (info.load_failed) > + DBUG_RETURN(ENCRYPTION_KEY_VERSION_INVALID); > + if (*buflen < info.length) > + { > + *buflen= info.length; > + DBUG_RETURN(ENCRYPTION_KEY_BUFFER_TOO_SMALL); > + } > + *buflen= info.length; > + memcpy(dstbuf, info.data, info.length); > + DBUG_RETURN(0); > +} > + > + > +/* Plugin defs */ > +struct st_mariadb_encryption aws_key_management_plugin= { > + MariaDB_ENCRYPTION_INTERFACE_VERSION, > + get_latest_key_version, > + get_key, > + // use default encrypt/decrypt functions > + 0, 0, 0, 0, 0 > +}; > + > + > +static TYPELIB key_spec_typelib = > +{ > + array_elements(key_spec_names) - 1, "", > + key_spec_names, NULL > +}; > + > +const char *log_level_names[] = > +{ > + "Off", > + "Fatal", > + "Error", > + "Warn", > + "Info", > + "Debug", > + "Trace", > + 0 > +}; > + > +static TYPELIB log_level_typelib = > +{ > + array_elements(log_level_names) - 1, "", > + log_level_names, NULL > +}; > + > +static MYSQL_SYSVAR_STR(master_key_id, master_key_id, > + PLUGIN_VAR_RQCMDARG | PLUGIN_VAR_MEMALLOC, > + "Key id for master encryption key. Used to create new datakeys. If not set, no new keys will be created", > + NULL, NULL, ""); > + > +static MYSQL_SYSVAR_ENUM(key_spec, key_spec, > + PLUGIN_VAR_RQCMDARG, > + "Encryption algorithm used to create new keys.", > + NULL, NULL, 0, &key_spec_typelib); > + > + > +static MYSQL_SYSVAR_ENUM(log_level, log_level, > + PLUGIN_VAR_RQCMDARG, > + "Logging for AWS API", > + NULL, NULL, 0, &log_level_typelib); > + > + > +static MYSQL_SYSVAR_INT(rotate_key, rotate_key, > + PLUGIN_VAR_RQCMDARG, > + "Set this variable to key id to perform rotation of the key. Specify -1 to rotate all keys", > + NULL, update_rotate, 0, -1, INT_MAX, 1); > + > +static struct st_mysql_sys_var* settings[]= { > + MYSQL_SYSVAR(master_key_id), > + MYSQL_SYSVAR(key_spec), > + MYSQL_SYSVAR(rotate_key), > + MYSQL_SYSVAR(log_level), > + NULL > +}; > + > +/* > + Plugin library descriptor > +*/ > +maria_declare_plugin(aws_key_management) > +{ > + MariaDB_ENCRYPTION_PLUGIN, > + &aws_key_management_plugin, > + "aws_key_management", > + "MariaDB Corporation", > + "AWS key management plugin", > + PLUGIN_LICENSE_GPL, > + plugin_init, > + plugin_deinit, > + 0x0100, > + NULL, > + settings, > + "1.0", > + MariaDB_PLUGIN_MATURITY_EXPERIMENTAL > +} > +maria_declare_plugin_end; Regards, Sergei Chief Architect MariaDB and security(a)mariadb.org

2 2

[Maria-developers] GSoC 2016 MHA GTID support.
by Charles Dirk 11 Mar '16

11 Mar '16

Hi, I am a computer science student at the university of Cape Town and wish to participate in GSoC 2016 with MariaDB Foundation. I am interested if working on making MHA understand GTIDs. I am proficient in C/C+, Python/Perl and Im equally familiar with replication. I look forward to hearing from mentors soon. Cheers, Charles.

3 3

[Maria-developers] MDEV-9709 Unexpected modification of value and warning about out of range value upon ALTER
by Alexander Barkov 11 Mar '16

11 Mar '16

Hi Elena, can you please review a patch for MDEV-9709? I'm quite sure about the code change, it's trivial. Please have a look into the tests. Thanks.

1 0