- developers - lists.mariadb.org

[Maria-developers] Updated Gtid_slave_pos of Untracked domain creates skipped events
by Sachin Setiya 23 Jul '18

23 Jul '18

Hi Andrei, Kristian! Problem:- This issue is regarding Mdev-9107 , where we have 3 master master with do_domain_ids of replication channel master domain-id . , When one replication channel is stopped and that master is updated , slave server updates the gtid slave_pos from second replication channel. And then when replication channel is brought up again then since we have updated the gtid_slave_pos we skip that event. I have created a abstract test case for this probem. (3) ^ ^ / \ (2)----->(1) 3 and 1 is configured with log slave updates. 3 has 2 replication channel m2_s3(do_domain_id=2 ), m1_s3(do_domain_id=1) 1 has one replication channel m2_s1(do_domain_id=2) I have tried with master_use_gtid == slave_pos and current_pos both give same result. When I stop replication channel m2_s3 and then write one-one event in (2) and (1) server. Server 3 will update its gtid_slave_pos and it will have latest gtid event seq no from domain id 2 and 1. When replication channel m2_s3 is bought up again since gtid_slave_pos of (3) is equal to (2) so we actually skipped the event generated on server 2. (I believe network lag between 2-->3 will have similar effect) (I have attached the test cases) Possible Solution. May be we should have one more option in master_use_gtid = binlog_state ? which will compare its binlog_state to arriving event gtid no ? (It is different from current_pos , current_pos will select the maximum seq_no between binlog_state and slave_pos) -- Regards Sachin Setiya Software Engineer at MariaDB

2 5

[Maria-developers] Please review a patch for MDEV-15758 Split Item_bool_func::get_mm_leaf() into virtual methods in Field and Type_handler
by Alexander Barkov 20 Jul '18

20 Jul '18

Hello Vicentiu, Please review my patch for MDEV-15758. It also fixes this bug: MDEV-15759 Expect "Impossible WHERE" for indexed_int_column=out_of_range_int_constant Thanks!

3 4

[Maria-developers] To MDEV-6608 Slave loses master binlog filename
by andrei.elkin＠pp.inet.fi 13 Jul '18

13 Jul '18

Hello, Kristian. I read through your analysis of the case to agree with the plan and its motivation: MDEV-6608> Generally, it is a very bad idea to destroy data during crash recovery ... try an approach that leaves the crashed binlog file intact, and instead ensures that the crashed binlog can be properly handled by the rest of the code. So in case of a partially written event group (transaction) the short (xid-less therefore) transaction is to be rolled back on the master. So why won't we properly enclose the group in the binlog at the end of recovery which means to append ROLLBACK Query-log-event, naturally before that the last event of the group needs sanity check? Was that that you had in mind actually? Cheers, Andrei

2 1

[Maria-developers] MDEV-13118 Wrong results with LOWER and UPPER and subquery
by Alexander Barkov 12 Jul '18

12 Jul '18

Hello Sergei, Please review a patch for MDEV-13118. Thanks!

2 4

[Maria-developers] Selecting Blob by parts (suppose select 10 to 100 char from blob )
by Sachin Setiya 12 Jul '18

12 Jul '18

Hello everyone! On optimizer call (11-7-2018) it was discussed , instead of selecting the whole blob , sometime we just need to select the subpart of it. And we should extend handler/engine api to facilitate this. Although this looks like a interesting idea , Serg (and I ) did not understand why we need this , what is use case of it. Some use case which I think can be like optimizing substring SELECT SUBSTRING(blob_col, 4,4) FROM t1; And we can push down this SUBSTRING condition directly to storage engine. But apart from this what use case this have ? -- Regards Sachin Setiya Software Engineer at MariaDB

1 0

[Maria-developers] What can MariaDB developers learn from Debian and Ubuntu Quality Assurance systems?
by Otto Kekäläinen 11 Jul '18

11 Jul '18

Hello! I had the presentation with the title in the subject line at the developers unconference some weeks ago. The slides are available here: https://mariadb.org/wp-content/uploads/2018/07/2018-06-27-mariadb-dev-learn… I recommend all developers skim though them. Debian has some really nice QA systems. All presentations from the event can be found at: https://mariadb.org/tampere-mariadb-developers-unconference-reportback-pres…

1 0

Re: [Maria-developers] 1a2748c: MDEV-15473 Isolate/sandbox PAM modules, so that they can't crash the server.
by Sergei Golubchik 10 Jul '18

10 Jul '18

Hi, Alexey! On Jul 10, Alexey Botchkov wrote: > > > Did you test that it works? > Of course i tested the mysql_install_db script a lot. > Didn't test the .rpm and .deb though. As far as i can see they plainly run > mysql_install_db --user=.. under root. could you please test with rpm and deb too? should only take 10-20 minutes in a buildbot VM. ok to push after that (if all goes well :)) Regards, Sergei Chief Architect MariaDB and security(a)mariadb.org

1 0

Re: [Maria-developers] 1a2748c: MDEV-15473 Isolate/sandbox PAM modules, so that they can't crash the server.
by Sergei Golubchik 09 Jul '18

09 Jul '18

Hi, Alexey! On Jul 09, Alexey Botchkov wrote: > revision-id: 1a2748ceb0e30770eebc97d82d7baa885b332c49 (mariadb-10.3.6-45-g1a2748c) > parent(s): aa01f51bdef9cc38d8e0a75ea9e2788651e41d16 > committer: Alexey Botchkov > timestamp: 2018-07-09 13:20:42 +0400 > message: > > MDEV-15473 Isolate/sandbox PAM modules, so that they can't crash the server. > > Scripts added to set safe permissions for the auth_pam_tool and it's > directory. > > diff --git a/scripts/mysql_install_db.sh b/scripts/mysql_install_db.sh > index ad7c028..fd26228 100644 > --- a/scripts/mysql_install_db.sh > +++ b/scripts/mysql_install_db.sh > @@ -308,6 +308,7 @@ then > srcpkgdatadir="$srcdir/scripts" > buildpkgdatadir="$builddir/scripts" > plugindir="$builddir/plugin/auth_socket" > + pamtooldir="$builddir/plugin/auth_pam" > elif test -n "$basedir" > then > bindir="$basedir/bin" # only used in the help text > @@ -337,6 +338,7 @@ then > exit 1 > fi > plugindir=`find_in_dirs --dir auth_socket.so $basedir/lib*/plugin $basedir/lib*/mysql/plugin` > + pamtooldir=`find_in_dirs --dir auth_pam.so $basedir/lib*/plugin $basedir/lib*/mysql/plugin` Why not just pamtooldir=$plugindir ? > else > basedir="@prefix@" > bindir="@bindir@" > @@ -445,6 +448,13 @@ done > > if test -n "$user" > then > + chown $user "$pamtooldir/auth_pam_tool_dir" > + if test $? -ne 0 > + then > + echo "Cannot change ownership of the '$pamtooldir\auth_pam_tool_dir' directory" > + echo " to the '$user' user. Check that you have the necessary permissions and try again." > + exit 1 > + fi Looks ok. Did you test that it works? I mean, a test woud be to install from a tarball and run mysql_install_db or install from rpm or deb (they run the install script automatically) and then check that all permissions are correct. Also, I suspect that at additional chown root "$pamtooldir/auth_pam_tool_dir/auth_pam_tool could also be useful here. But not when the script is running in the builddir. Regards, Sergei Chief Architect MariaDB and security(a)mariadb.org

1 0

Re: [Maria-developers] [Commits] 765ba2a: MDEV-15473 Isolate/sandbox PAM modules, so that they can't crash the server.
by Sergei Golubchik 06 Jul '18

06 Jul '18

Hi, Alexey! On Jul 05, Alexey Botchkov wrote: > revision-id: 765ba2ac76dab984183bf829dc3407713a4d5d9b (mariadb-10.3.6-40-g765ba2a) > parent(s): 7e704a2308e25953b5f8fb154eb325df3e25c2ec > committer: Alexey Botchkov > timestamp: 2018-07-05 17:00:47 +0400 > message: > > MDEV-15473 Isolate/sandbox PAM modules, so that they can't crash the server. > > Proper access permissions for the auth_pam_tool_dir and auth_pam_tool. > > diff --git a/plugin/auth_pam/CMakeLists.txt b/plugin/auth_pam/CMakeLists.txt > index 4943d57..b9313de 100644 > --- a/plugin/auth_pam/CMakeLists.txt > +++ b/plugin/auth_pam/CMakeLists.txt > @@ -11,7 +11,13 @@ IF(HAVE_PAM_APPL_H) > ADD_DEFINITIONS(-D_GNU_SOURCE) > MYSQL_ADD_PLUGIN(auth_pam_v1 auth_pam_v1.c LINK_LIBRARIES pam MODULE_ONLY) > MYSQL_ADD_PLUGIN(auth_pam auth_pam.c LINK_LIBRARIES pam dl MODULE_ONLY) > - MYSQL_ADD_EXECUTABLE(auth_pam_tool auth_pam_tool.c DESTINATION ${INSTALL_PLUGINDIR}/auth_pam_tool_dir COMPONENT Server) > + MYSQL_ADD_EXECUTABLE(auth_pam_tool auth_pam_tool.c DESTINATION ${INSTALL_PLUGINDIR}/auth_pam_tool_dir COMPONENT Server) > TARGET_LINK_LIBRARIES(auth_pam_tool pam) > + INSTALL(CODE "EXECUTE_PROCESS( > + COMMAND chmod u=rwx,g=,o= auth_pam_tool_dir > + COMMAND chmod u=rwx,g=rx,o=rx auth_pam_tool_dir/auth_pam_tool > + COMMAND chmod +s auth_pam_tool_dir/auth_pam_tool > + WORKING_DIRECTORY \$ENV{DESTDIR}\${CMAKE_INSTALL_PREFIX}/${INSTALL_PLUGINDIR}/)" > + COMPONENT Server) I think it's generally ok. Two comments: minor: you can combine two chmods on auth_pam_tool in one, like u=rwxs major: you still need to make auth_pam_tool_dir to be owned by mysql user. I'm afraid the only way to do it is from a post-install scriptlet or from mysql_install_db. Regards, Sergei Chief Architect MariaDB and security(a)mariadb.org

1 0

Re: [Maria-developers] [Commits] af15686: MDEV-16242: MyRocks: parallel slave on a table without PK can stop with ER_KEY_NOT_FOUND
by Kristian Nielsen 05 Jul '18

05 Jul '18

psergey(a)askmonty.org (Sergei Petrunia) writes: > MDEV-16242: MyRocks: parallel slave on a table without PK can stop with ER_KEY_NOT_FOUND > > DRAFT: If RBR event applier uses a secondary key or a full table scan > to locate a row, force waiting for prior commit to complete. I think if you wait_for_prior_commit() also for secondary key case, you should definitely restrict it to be only for conservative parallel replication, not for optimistic/agressive. Using a secondary key seems generally fine (as opposed to full table scan), and this problem should affect only conservative mode, no need to penalise optimistic/aggressive. (But maybe that is what is meant with "DRAFT"). - Kristian.

2 2