March 2015 - developers - lists.mariadb.org

[Maria-developers] MariaDB Maxscale Project (GSOC 2015)
by Aalekh Nigam 09 Mar '15

09 Mar '15

Hello, I am Aalekh Nigam, 3rd year Undergraduate Student at JIIT, Noida. I have recently successfully completed Mozilla Winter Of Security and Owasp Winter of Code, and is aiming for MaxScale projects for this GSOC. Can someone please guide me through starter’s bug for Maxscale to get my hands on with MaxScale development ? Thank You —Aalekh

2 1

[Maria-developers] MDEV-7643 review
by Elena Stepanova 09 Mar '15

09 Mar '15

Hi Sergei, http://lists.askmonty.org/pipermail/commits/2015-March/007508.html Could you please check patch for https://mariadb.atlassian.net/browse/MDEV-7643 (MTR creates nested links when tests are run with --mem)? It's in the area that you looked at recently, so you might think of something that I could have missed. Thanks, Elena

2 1

Re: [Maria-developers] [Commits] e464bc6: MDEV-7107 Sporadic test failure in multi_source.multisource
by Kristian Nielsen 09 Mar '15

09 Mar '15

elenst(a)montyprogram.com writes: > revision-id: e464bc6fe3d7566831da5e26447d90e758b5aa2a > committer: Elena Stepanova > timestamp: 2015-03-09 02:07:47 +0200 > message: > > MDEV-7107 Sporadic test failure in multi_source.multisource > > Extend show_slave_status.inc to run SHOW ALL SLAVES STATUS and > SHOW SLAVE 'name' STATUS on demand, and make the test use > the include file instead of direct SHOW statements Looks fine, thanks! Ok to push. I agree with your changes to show_slave_status.inc. I think they will be useful for futute test cases also. - Kristian.

1 0

[Maria-developers] MDEV-7107 review
by Elena Stepanova 09 Mar '15

09 Mar '15

Hi Kristian, Could you please take a quick look at the patch for MDEV-7107 (Sporadic test failure in multi_source.multisource). http://lists.askmonty.org/pipermail/commits/2015-March/007543.html I don't expect you'll mind changes to the test itself, we discussed earlier that's how SHOW SLAVE STATUS should be handled; but please check if you're okay with the additions to show_slave_status.inc, or you would prefer a separate include file for SHOW ALL SLAVES STATUS instead. Thanks, Elena

1 0

[Maria-developers] rpl test with gtid as default
by Jonas Oreland 09 Mar '15

09 Mar '15

what do you think ? /Jonas

3 3

Re: [Maria-developers] a1c4620: MDEV-6858: enforce_storage_engine option
by Sergei Golubchik 09 Mar '15

09 Mar '15

Hi, Jan! On Mar 08, Jan Lindström wrote: > revision-id: a1c4620646d43eaff979c132a57ee2101070fa5e > parent(s): 1b74f32b1d94465f6d6e14e4aa7ba1f94c32e39e > committer: Jan Lindström > branch nick: 10.1-innodb > timestamp: 2015-03-08 17:23:05 +0200 > message: > > MDEV-6858: enforce_storage_engine option > > Merge from Percona Server enforced use of a specific storage engine > authored by Stewart Smith. > > Modified to be session variable and modifiable only by SUPER. Use > similar implementation as default_storage_engine. > > diff --git a/mysql-test/t/enforce_storage_engine.test b/mysql-test/t/enforce_storage_engine.test > new file mode 100644 > index 0000000..36231e6 > --- /dev/null > +++ b/mysql-test/t/enforce_storage_engine.test > @@ -0,0 +1,35 @@ ... > +--error ER_UNKNOWN_STORAGE_ENGINE > +CREATE TABLE t2 (c1 INT PRIMARY KEY AUTO_INCREMENT, c2 VARCHAR(10)) ENGINE=MyISAM; > + > +SET SESSION enforce_storage_engine=MyISAM; > +select @@session.enforce_storage_engine; > +select * from t1; > +drop table t1; > + > +--error 1286 > +SET SESSION enforce_storage_engine=FooBar; > + > +select @@session.enforce_storage_engine; Consider adding tests for "only modifiable by SUPER" (create a new user, connect with it, see that enforce_storage_engine cannot be changed), and for modified enforce_storage_engine (you tested that it can be changed, but didn't do any CREATE TABLE with the new value). > diff --git a/sql/handler.cc b/sql/handler.cc > index 12d7ffb..5d5ac30 100644 > --- a/sql/handler.cc > +++ b/sql/handler.cc > @@ -117,6 +117,15 @@ static plugin_ref ha_default_tmp_plugin(THD *thd) > return ha_default_plugin(thd); > } > > + > +static plugin_ref ha_enforced_plugin(THD *thd) > +{ > + if (thd->variables.enforced_table_plugin) > + return thd->variables.enforced_table_plugin; > + return NULL; > +} > + > + > /** @brief > Return the default storage engine handlerton for thread > > @@ -148,6 +157,25 @@ handlerton *ha_default_tmp_handlerton(THD *thd) > > > /** @brief > + Return the enforced storage engine handlerton for thread > + > + SYNOPSIS > + ha_enforce_handlerton(thd) > + thd current thread > + > + RETURN > + pointer to handlerton OR NULL > + > +*/ > +handlerton *ha_enforced_handlerton(THD *thd) > +{ > + plugin_ref plugin= ha_enforced_plugin(thd); > + if (plugin) > + return plugin_hton(plugin); > + return NULL; > +} > + > +/** @brief > Return the storage engine handlerton for the supplied name > > SYNOPSIS > diff --git a/sql/mysqld.cc b/sql/mysqld.cc > index b97742d..6966fd9 100644 > --- a/sql/mysqld.cc > +++ b/sql/mysqld.cc > @@ -4738,8 +4738,8 @@ static void add_file_to_crash_report(char *file) > #define init_default_storage_engine(X,Y) \ > init_default_storage_engine_impl(#X, X, &global_system_variables.Y) > > -static int init_default_storage_engine_impl(const char *opt_name, > - char *engine_name, plugin_ref *res) > +int init_default_storage_engine_impl(const char *opt_name, > + char *engine_name, plugin_ref *res) You've made init_default_storage_engine_impl() extern, but you never use it anywhere. And you've forgot to create a command line option for enforced_table_plugin. > { > if (!engine_name) > { > diff --git a/sql/sql_table.cc b/sql/sql_table.cc > index e9a1ae9..184c073 100644 > --- a/sql/sql_table.cc > +++ b/sql/sql_table.cc > @@ -9745,12 +9745,24 @@ static bool check_engine(THD *thd, const char *db_name, > DBUG_ENTER("check_engine"); > handlerton **new_engine= &create_info->db_type; > handlerton *req_engine= *new_engine; > + handlerton *enf_engine= ha_enforced_handlerton(thd); I'd write it directly as thd->variables.enforced_table_plugin ? plugin_hton(thd->variables.enforced_table_plugin) : NULL; Or as an inlined function here, in sql_table.cc. Not as two small but non-inlined functions in handler.cc, that are only used once here. > bool no_substitution= thd->variables.sql_mode & MODE_NO_ENGINE_SUBSTITUTION; > *new_engine= ha_checktype(thd, req_engine, no_substitution); > DBUG_ASSERT(*new_engine); > if (!*new_engine) > DBUG_RETURN(true); > > + if (enf_engine) > + { > + if (enf_engine != *new_engine && no_substitution) > + { > + const char *engine_name= ha_resolve_storage_engine_name(req_engine); > + my_error(ER_UNKNOWN_STORAGE_ENGINE, MYF(0), engine_name, engine_name); > + DBUG_RETURN(TRUE); > + } > + *new_engine= enf_engine; > + } > + > if (req_engine && req_engine != *new_engine) > { > push_warning_printf(thd, Sql_condition::WARN_LEVEL_NOTE, > diff --git a/sql/sys_vars.cc b/sql/sys_vars.cc > index 8371df0..cd53714 100644 > --- a/sql/sys_vars.cc > +++ b/sql/sys_vars.cc > @@ -3437,6 +3437,28 @@ static Sys_var_plugin Sys_default_tmp_storage_engine( > SESSION_VAR(tmp_table_plugin), NO_CMD_LINE, > MYSQL_STORAGE_ENGINE_PLUGIN, DEFAULT(&default_tmp_storage_engine)); > > +extern int init_default_storage_engine_impl(const char* opt_name, char *engine_name, plugin_ref *plugin); Not used below. > + > + > +static bool check_super_and_not_null(sys_var *self, THD*thd, set_var *var) I think that NULL value should be allowed, it means "don't enforce a storage engine". > +{ > +#ifndef NO_EMBEDDED_ACCESS_CHECKS > + if (!(thd->security_ctx->master_access & SUPER_ACL)) > + { > + my_error(ER_SPECIFIC_ACCESS_DENIED_ERROR, MYF(0), "SUPER"); > + return true; > + } > +#endif > + return var->value && var->value->is_null(); > +} > + > +static Sys_var_plugin Sys_enforce_storage_engine( > + "enforce_storage_engine", "Force the use of a storage engine for new " > + "tables", > + SESSION_VAR(enforced_table_plugin), > + NO_CMD_LINE, MYSQL_STORAGE_ENGINE_PLUGIN, > + DEFAULT(0), NO_MUTEX_GUARD, NOT_IN_BINLOG, ON_CHECK(check_super_and_not_null)); > + > #if defined(ENABLED_DEBUG_SYNC) > /* > Variable can be set for the session only. Regards, Sergei

1 0

[Maria-developers] [GSoC'15] MaxScale Project
by sriram patil 07 Mar '15

07 Mar '15

Hi Markus, I am Sriram, a Masters' student at IIIT Hyderabad, India. I have worked with MariaDB in GSoC-2014 on MDEV-5359 under Alexander Barkov. I have also documented my work on my blog <http://srirampatil.tumblr.com/>. I am interested to work on "MaxScale filter to capture incoming operations for consumption in external sources". I have downloaded and built MariaDB 10.1 source code from github. I was trying to build MaxScale but it resulted in some errors. I followed the installation instructions given on MaxScale github(develop branch). The error is as follows /home/sriram/gsoc/MaxScale/query_classifier/query_classifier.cc: In function ‘bool skygw_stmt_causes_implicit_commit(LEX*, int*)’: /home/sriram/gsoc/MaxScale/query_classifier/query_classifier.cc:848:40: error: ‘struct LEX’ has no member named ‘drop_temporary’ succp = !(lex->drop_temporary); ^ make[2]: *** [query_classifier/CMakeFiles/query_classifier.dir/query_classifier.cc.o] Error 1 make[1]: *** [query_classifier/CMakeFiles/query_classifier.dir/all] Error 2 I checked struct LEX from MariaDB source code and it does not have the mentioned field. So, is MaxScale expecting some other MariaDB source code? It would be great if you could give me some pointers regarding the same. Thanks, Sriram Patil

2 2

Re: [Maria-developers] GSoC 15 : MaxScale Project under Markus Makela
by Markus Mäkelä 06 Mar '15

06 Mar '15

Hi, The most up to date documentation for MaxScale can be found on the develop branch on GitHub: https://github.com/mariadb-corporation/MaxScale/blob/develop/Documentation/… I suggest using that rather than the PDF documentation since it's a lot more easy to use. The filter interface can be found in the server/include/filter.h file. All filters must implement these functions for them to work with MaxScale. The filters are located in the server/modules/filter folder. The best way to get a hang of the filters is to take a look at one and the regexfilter and qlafilter are the ones that show the filter interface in action without having much code in them. Other useful stuff to look at can be found in the server/include folder are buffer.h(MaxScale uses these to route queries though it), modutils.h(useful functions for SQL handling), housekeeper.h(for tasks that take a long time this can be used) and hashtable.h(a generic hashtable). All of the MaxScale team works around UTC+1. For quick questions ask them on #maxscale on Freenode but email will most likely be answered as fast as IRC. Markus > Hello guys, > > As per the official list > <https://mariadb.com/kb/en/mariadb/google-summer-of-code-2015/> of projects > by MariaDB, I'm interested in the project - MaxScale filter to capture > incoming operations for consumption in external sources ( Mentor : Markus > Makela ) > > *I've 3 requests/doubts : * > > 0.I've downloaded the MaxScale code from github > <https://github.com/mariadb-corporation/MaxScale> and have read the > documentation. > It says the filter configuration file is : MaxScale.ini > I didn't find this file in the codebase. > > Nor was I able to figure out the code-location related to the "filter" > module. > > 1.Kindly suggest for the further things I should do/try. > > Shall I install MariaDB & MaxScale? > What part of code-base should I look into & play around? > > 2.What is the UTC time when most of you guys are online on IRC-channel > (#maria)? > > Since I'm from India, I need to know the time when most of you guys are > online to get immediate help on IRC. > > > Thanks in advance. > > -- > Lokesh Walase -- Markus Mäkelä, Software Engineer MariaDB Corporation t: +358 40 7740484 | Skype: markus.j.makela

1 0

[Maria-developers] Comments/thoughts on patch
by Jonas Oreland 06 Mar '15

06 Mar '15

Hi Kristian, I wrote attached patch (so that i would understand what code actually does) and found myself while refactoring also doing a change, namely that i removed the check for server_id in a couple of places. it still passes all tests... could you comment on this, or write(outline) a testcase that fails with my patch. i think functionality that i might have removed would be if you write have 2 mysqld M1 & M2 and your replicate both way...and you *don't* use domain_id...and you write concurrently on both mysqld (but maybe disjoint data ??). feedback/comments appreciated. /Jonas ps. patch applies to f4f37533a09b9776e8d5ac3f3a27957f553c9043

2 9

[Maria-developers] MaxScale JIRA project
by Vangelis Katsikaros 05 Mar '15

05 Mar '15

Hi While browsing the GSOC projects, I noticed that the MX project, and the tickets mentioned, require signing in. I don't know if this intentional or not (the MDEV project doesn't require signin for viewing the project). Just thought I should mention it. Regards Vangelis

2 2