June 2016 - developers - lists.mariadb.org

[Maria-developers] Executing query inside mysql server
by knizhnik 24 Jun '16

24 Jun '16

Hi, I wonder what is the simplest way to execute SQL query inside server? In PostgreSQL there is SPI API, SQLite allows to use client API,... but I failed to find right way of running queries inside mySQL server. My task is the following: I want to patch mySQL server by redirecting execution of some queries to some external executor. The text of this queries will be stored inside database in some special table. So when some query is executed by mySQL server, I want to perform lookup in this table. What is the simplest way to do it in mySQL?

2 1

[Maria-developers] Why/how does the MariaDB client depend on libdbd-mysql-perl?
by Otto Kekäläinen 24 Jun '16

24 Jun '16

I am returning to this age-old discussion about libdbd-mysql-perl as dependency in Debian: 2015-07-21 19:50 GMT+03:00 Sergei Golubchik <vuvova(a)gmail.com>: >> + libdbi-perl, >> + libterm-readkey-perl, >> + mariadb-client-core-10.0 (>= ${source:Version}), >> + ${misc:Depends}, >> + ${perl:Depends}, >> + ${shlibs:Depends} >> +Recommends: libdbd-mysql-perl (>= 1.2202) > > really? why libdbi-perl is required but libdbd-perl is optional? > the one makes no sense without the other. and libterm-readkey-perl > is required, why it's really optional, everything works without it. In MySQL 5.6 this dependency also still exists: http://anonscm.debian.org/cgit/pkg-mysql/mysql-5.6.git/tree/debian/control Package: mysql-client-5.6 Architecture: any Depends: debianutils (>=1.6), libdbd-mysql-perl (>= 1.2202), libdbi-perl, libterm-readkey-perl, But it has been dropped in MySQL 5.7 packaging. In 5.7 it is not even a Recommends or Suggests, but completely removed. http://anonscm.debian.org/cgit/pkg-mysql/mysql-5.7.git/tree/debian/control I don't know what to do about this because I don't really understand what or how the mysql client does with perl library dependencies. It is a C program, does it really use the perl library in any way? So auxilary script does something? Can we drop the dependency on libdbd-mysql-perl, and starting from what version? - Otto

2 1

[Maria-developers] MDEV-8502 DECIMAL accepts out of range DEFAULT values
by Alexander Barkov 24 Jun '16

24 Jun '16

Hello Sergei, Please review a patch fixing: MDEV-8502 DECIMAL accepts out of range DEFAULT values MDEV-10277 Redundant NOTE when inserting '0.00001 ' into a DECIMAL(2,1) column Thanks.

1 0

Re: [Maria-developers] [MariaDB/server] MDEV-10271: Log stop of slave SQL thread on errors, too (#188)
by Kristian Nielsen 23 Jun '16

23 Jun '16

Hartmut Holzgraefe <notifications(a)github.com> writes: > -- Commit Summary -- > > * MDEV-10271: Log stop of slave SQL thread on errors, too > * MDEV-10272: add master host/port info to slave thread exit messages So one comment. With the first patch, there are two cases where the error message "Slave SQL thread exiting, replication stopped ..." will be logged even though the thread fails before it gets to log the corresponding start message "Slave SQL thread initialized, starting replication ..." Was this intended, it seems perhaps a bit confusing if the log has spurious messages about threads stopping, when there are no corresponding start messages? If not intended, it is simple to change by adding an extra label err_after_start: before the message, and using that or err: as appropriate in the gotos. What do you think? But I see you filed pull requests against 10.0 and 10.1, so it seems you were intending these for the stable releases? I do not have any strong opinion on this myself, but for stable releases, you should write an email to maria-developers(a)lists.launchpad.net to solicit opinions about making an exception for this case. Include justification of why it should go in stable releases (and not just in 10.2), and discussion of possible impact on existing applications/setups. If there are no objections, I'll push them. - Kristian.

1 0

[Maria-developers] Formatting rules for double => varchar assignment
by Jesse Schalken 23 Jun '16

23 Jun '16

Hi MariaDB Developers, I have an application which depends on the formatting MariaDB applies to a double when assigned to a char/varchar column of different lengths, and am trying to reproduce the behaviour in PHP. I have gotten some way but just can't seem to get the rules right. I'm having trouble finding a specification for this in the docs or the code that is responsible for this in the MariaDB source. Can someone point me in the right direction? Thanks

2 4

Re: [Maria-developers] GSoC weekly reports (Unique indexes for blobs)
by Shubham Barai 22 Jun '16

22 Jun '16

GSoC (week 4) Hello everyone, After working on create table operation,next thing I had to work on was insert operations.So I explored some of the functions like row_ins_scan_index_for_duplicates, btr_pcur_get_rec to get clear understanding about how to implement duplicate search on hash index. There was a problem in hash function that I wrote .It would calculate same hash value for two different keys if the prefix length of blob key part was zero. Now it seems to be working after I checked it in debugger.I still have to modify it for data types like varchar etc. I have added test cases for insert operations in myisam. In MyIsam, I found one problem in update operation. When updating a row,if the key is conflicting then server crashes because some pointer goes invalid in compare_record. I haven't fixed this issue yet. I also modified some functions in dict0load.cc to adjust some members of dict_index_t for a new index type.The main problem is that index entry for hash based index cointains only two fields(hash value and row id) while dict_index_t contains hash field and other user defined fields which are used to calculate hash value.Some of the operations like alter table( e.g. rename column) needs to get access to all fields while other functions like rec_get_offsets and row_build_index_entry_low needs to get access to only hash field and row id. I am still working on this to find efficient solution to this problem. On 16 June 2016 at 23:29, Sergei Golubchik <vuvova(a)gmail.com> wrote: > Hi, Shubham! > > What I wanted to say on IRC was: > > here's what the comment of cmp_dtuple_rec_with_match_low() says: > > ............... If rec has an externally stored field we do not > compare it but return with value 0 if such a comparison should be > made. > > Note that blobs are externally stored fields in InnoDB, so, I think, > this means that you cannot use cmp_dtuple_rec() to compare blobs. > > Regards, > Sergei > Chief Architect MariaDB > and security(a)mariadb.org >

2 1

Re: [Maria-developers] [Commits] 511bd2c: MDEV-10216: Assertion `strcmp(share->unique_file_name, filename) ||
by Nirbhay Choubey 21 Jun '16

21 Jun '16

Hi Monty, On Tue, Jun 21, 2016 at 9:30 AM, Michael Widenius <monty(a)mariadb.org> wrote: > > Hi! > > >>>>> "Nirbhay" == Nirbhay Choubey <nirbhay(a)mariadb.com> writes: > > <cut> > > Nirbhay> ALTER TABLE implementation takes a shortcut for operations not > affecting > Nirbhay> .frm files, > Nirbhay> which includes simple RENAMEs. But this is skipped for temp > tables and thus > Nirbhay> copying > Nirbhay> always takes place for temp tables. > > Nirbhay> The following commit (from 2002) added this exception for temp > tables: > Nirbhay> > https://github.com/MariaDB/server/commit/a7798dfd0a6472bf65fffc2ade543605e1… > > The above old patch is not really working for the current code. > > Nirbhay> I have created a patch to also include temp tables to this > shortcut : > Nirbhay> https://gist.github.com/nirbhayc/442a0c269ce48b283543cac434aaf44e > > This solves the problem of doing a simple rename, but not for things > like: > > alter table t1 add column c int, rename to t2; > > Which will still produce crashes. > > When it comes to last_version, is still vote for that we should be using > THD::rename_temporary_table when renaming temporary tables instead of > doing any handler calls. > Right. > > Maybe we could move THD::rename_temporary_tables to ha_rename_table to > get all temporary tables handled in one place ? > handler::ha_rename_table() is currently being used only for non-temporary tables at places which do not support temporary tables like partitions, DDL logs, RENAME TABLE, etc. So, moving rename_temporary_table to ha_rename_table seem overkill and possibly inappropriate. Here's the latest version of patch : http://lists.askmonty.org/pipermail/commits/2016-June/009477.html Best, Nirbhay > Regards, > Monty >

1 0

[Maria-developers] Different results from UNION and UNION DISTINCT
by Pantelis Theodosiou 21 Jun '16

21 Jun '16

According to this post at Stack Overflow, MariaDB 10.1.10 gives wrong results (duplicate rows) for UNION DISTINCT while the UNION version gives correct ones. (notice the comment: http://dba.stackexchange.com/questions/141799/select-distinct-values-of-two… ) The query in question is very simple: SELECT a AS result FROM tableXUNION DISTINCTSELECT b FROM tableX ; I couldn't find a related bug in the MDEV. Is this a known (or fixed) issue or should I go and add a bug report/ Pantelis Theodosiou

2 1

Re: [Maria-developers] 13f13b7: MDEV-5973: MySQL Bug#11757486:49539: NON-DESCRIPTIVE ERR (ERROR 0 FROM STORAGE ENGINE) WITH MULTI-TABLE UPDATE
by Sergei Golubchik 21 Jun '16

21 Jun '16

Hi, Oleksandr! On Jun 21, Oleksandr Byelkin wrote: > revision-id: 13f13b7b2bede46c90772a105a502c33fe93f4d1 (mariadb-10.0.25-11-g13f13b7) > parent(s): 70ad689b11bfbd8a30a777f4893a5384628c00e7 > committer: Oleksandr Byelkin > timestamp: 2016-06-21 18:37:22 +0200 > message: > > MDEV-5973: MySQL Bug#11757486:49539: NON-DESCRIPTIVE ERR (ERROR 0 FROM STORAGE ENGINE) WITH MULTI-TABLE UPDATE > > Condition in processing IGNORE clause for UPDATE & multi-table UPDATE made the same. ok to push, thanks! Regards, Sergei Chief Architect MariaDB and security(a)mariadb.org

1 0

[Maria-developers] MDEV-9524 Cannot load from mysql.event when sql_mode is set to PAD_CHAR_TO_FULL_LENGTH
by Alexander Barkov 21 Jun '16

21 Jun '16

Hello Sergei, Please review a patch for mdev-9524. Thanks.

2 3